DETAILED ACTION
This action is in reply to the Amendments filed on 11/03/2025.
Claims 1-20 are rejected.
Claims 1-20 are currently pending and have been examined.
Response to Amendment
Applicant’s amendment, filed 11/03/2025, has been entered. Claims 1, 7, 9, 11, 16, 18, and 20 have been amended.
Claim Objections
The claim objections from the prior Office Action have been withdrawn pursuant Applicant’s amendments.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Under Step 1 of the Subject Matter Eligibility Test for Products and Processes, the claims must be directed to one of the four statutory categories. All the claims are directed to one of the four statutory categories (YES).
Under Step 2A in MPEP 2106.04, it is determined whether the claims are directed to a judicially recognized exception. Step 2A is a two-prong inquiry.
Under Prong 1, it is determined whether the claim recites a judicial exception (YES). Taking Claim 20 as representative, the claim recites limitations that fall within the certain methods of organizing human activity groupings of abstract ideas, including:
-a memory storing instructions; and
-at least one processor configured to execute the instructions to perform operations comprising:
-retrieving, from one or more data structures: a product search query by a user comprising at least an alphanumeric product model number, a text string, or any combination thereof, at least one data set comprising at least a catalogue of product model numbers collected over a predefined time frame, and a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers;
-determining, using at least one machine-learning algorithm: a search type, and a plurality of attributes associated with the product query comprising at least a product model number, a product name, or product description;
-performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features;
-upon performing the iterative match, determining no matches between the product search query and the product index data set;
-based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query;
-transmitting the at least one top alternative product for display to the user on an interactive web page of a user device, the [display] web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product; and
-transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order
The above limitations recite the concept determining and providing an alternative product for purchase. The above limitations fall within the “Certain Methods of Organizing Human Activity” groupings of abstract ideas, enumerated in MPEP 2106.04(a).
Certain methods of organizing human activity include:
fundamental economic principles or practices (including hedging, insurance, and mitigating risk)
commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; and business relations)
managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions)
The limitations of performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features; and upon performing the iterative match, determining no matches between the product search query and the product index data set are processes that, under their broadest reasonable interpretation, cover a commercial interaction. For example, “performing” and “determining” in the context of this claim encompass advertising, and marketing or sales activities.
Similarly, the limitations of at least one processor configured to execute the instructions to perform operations comprising: retrieving, from one or more data structures: a product search query by a user comprising at least an alphanumeric product model number, a text string, or any combination thereof, at least one data set comprising at least a catalogue of product model numbers collected over a predefined time frame, and a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers; determining, using at least one machine-learning algorithm: a search type, and a plurality of attributes associated with the product query comprising at least a product model number, a product name, or product description; based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; transmitting the at least one top alternative product for display to the user on an interactive web page of a user device, the [display] web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product; and transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order are processes that, under their broadest reasonable interpretation, cover a commercial interaction. That is, other than reciting that the operations are performed by at least one processor configured to execute the instructions, that the retrieving is from one or more data structures, that the determining is using at least one machine-learning algorithm, that the determining of at least one top alternative product is using a machine-learning algorithm, that the display is on an interactive web page of a user device, that the displaying is on a web page, that the user elements are interactive user elements, that the selection from the user is on an interface of the web page, and that the transmitting of the selection is to the system, nothing in the claim element precludes the step from practically being performed by people. For example, but for the “at least one processor configured to execute the instructions,” “one or more data structures,” “at least one machine-learning algorithm,” “a machine learning algorithm,” “an interactive web page,” “a user device,” “interactive user elements,” “an interface of the web page,” and “the system” language, “retrieving,” “determining,” “determine,” “transmitting,” and “transmitting” in the context of this claim encompasses advertising, and marketing or sales activities.
Under Prong 2, it is determined whether the claim recites additional elements that integrate the exception into a practical application of the exception. This judicial exception is not integrated into a practical application (NO).
-a memory storing instructions; and
-at least one processor configured to execute the instructions to perform operations comprising:
-retrieving, from one or more data structures: a product search query by a user comprising at least an alphanumeric product model number, a text string, or any combination thereof, at least one data set comprising at least a catalogue of product model numbers collected over a predefined time frame, and a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers;
-determining, using at least one machine-learning algorithm: a search type, and a plurality of attributes associated with the product query comprising at least a product model number, a product name, or product description;
-performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features;
-upon performing the iterative match, determining no matches between the product search query and the product index data set;
-based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query;
-transmitting the at least one top alternative product for display to the user on an interactive web page of a user device, the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product; and
-transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order
These limitations are not indicative of integration into a practical application because:
The additional elements of claim 20 are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than mere instructions to implement or apply the abstract idea on a generic computing hardware (or, merely use a computer as a tool to perform an abstract idea) as supported by paragraph [0090] of Applicant’s specification – “Various programs or program modules can be created using any of the techniques known to one skilled in the art or can be designed in connection with existing software.” Specifically, the additional elements of a computer-implemented system, a memory storing instructions, at least one processor configured to execute the instructions, one or more data structures, at least one machine-learning algorithm, a machine learning algorithm, an interactive web page, a user device, interactive user elements, an interface of the web page, and the system are recited at a high-level of generality (i.e. as a generic processor performing the generic computer functions of retrieving data, determining data, performing a match, and transmitting data) such that they amount do no more than mere instructions to apply the exception using generic computer components. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea. Further, the additional elements do no more than generally link the use of the judicial exception to a particular technological environment or field of use (such as computers or computing networks). Employing well-known computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment, does not integrate the exception into a practical application.
Additionally, the additional elements are insufficient to integrate the abstract idea into a practical application because the claim fails to i) reflect an improvement in the functioning of a computer or an improvement to another technology or technical field, ii) apply the judicial exception with, or use the judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, iii) effect a transformation or reduction of a particular article to a different state or thing, or iv) apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment.
Accordingly, the judicial exception is not integrated into a practical application.
Under Step 2B, it is determined whether the claims recite additional elements that amount to significantly more than the judicial exception. The claims of the present application do not include additional elements that are sufficient to amount to significantly more than the judicial exception (NO).
In the case of claim 20, taken individually or as a whole, the additional elements of claim 21 do not provide an inventive concept. As discussed above under step 2A (prong 2) with respect to the integration of the abstract idea into a practical application, the additional elements used to perform the claimed functions amount to no more than a general link to a technological environment.
Even considered as an ordered combination (as a whole), the additional elements do not add anything significantly more than when considered individually.
Claim 1 is a computer-implemented system reciting similar functions as claim 20. Examiner notes that claim 1 recites the additional elements of a computer-implemented system, a memory storing instructions, at least one processor configured to execute the instructions, one or more data structures, at least one machine-learning algorithm, a machine learning algorithm, an interactive web page, a user device, interactive user elements, an interface of the web page, and the system, claim 1 does not qualify as eligible subject matter for similar reasons as claim 20 indicated above.
Claim 11 is a computer-implemented method reciting similar functions as claim 20. Examiner notes that claim 1 recites the additional elements of a computer-implemented method, one or more data structures, at least one machine-learning algorithm, a machine learning algorithm, an interactive web page, a user device, interactive user elements, an interface of the web page, and the system, however, claim 1 does not qualify as eligible subject matter for similar reasons as claim 20 indicated above.
Therefore, claims 1 and 11 do not provide an inventive concept and do not qualify as eligible subject matter.
Dependent claims 2-10 and 12-19, when analyzed as a whole, are held to be patent ineligible under 35 U.S.C. § 101 because they do not add “significantly more” to the abstract idea. More specifically, dependent claims 2-10 and 12-19 further fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas in that they recite commercial interactions. Dependent claims 2-4, 6-10, 12-13, and 15-19, do not recite any farther additional elements, and as such are not indicative of integration into a practical application for at least similar reasons discussed above. Dependent claims 5 and 14 recite the additional elements of the data structures, linear data structures, and non-linear data structures, but similar to the analysis under prong two of Step 2A these additional elements are used as a tool to perform the abstract idea. As such, under prong two of Step 2A, claims 2-10 and 12-19 are not indicative of integration into a practical application for at least similar reasons as discussed above. Thus, dependent claims 2-10 and 12-19 are “directed to” an abstract idea. Next, under Step 2B, similar to the analysis of claims 1, 11 and 20, dependent claims 2-10 and 12-19 when analyzed individually and as an ordered combination, merely further define the commonplace business method (i.e. determining and providing an alternative product for purchase) being applied on a general-purpose computer and, therefore, do not amount to significantly more than the abstract idea itself. Accordingly, the Examiner concludes that there are no meaningful limitations in the claims that transform the judicial exception into a patent eligible application such that the claims amount to significantly more than the judicial exception itself. The analysis above applies to all statutory categories of invention.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-3, 5-6, 11-12, 14-15, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al. (US 2021/0358007 A1), hereinafter Zhang, in view of Sun et al. (US 2011/0078157 A1), newly cited and hereinafter Sun, in further view of Garg et al. (US 9,830,392 B1), hereinafter Garg.
Regarding claim 1, Zhang discloses a computer-implemented system for identifying a target product and generating alternative product recommendations based on a user query, the system comprising:
-a memory storing instructions (Zhang, see at least: “The storage 710 may use any suitable combination of any suitable volatile and non-volatile physical storage mediums, including, for example, hard disk drives, solid state drives, optical media, flash memory, tape drives, registers, and random access memory, or the like, or any combination thereof” [0047]); and
-at least one processor configured to execute the instructions to perform operations (Zhang, see at least: “When implemented on a general-purpose microprocessor, the computer program code segments configure the microprocessor to create specific logic circuits” [0061]) comprising:
-retrieving, from one or more data structures (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048]):
-a product search query by a user (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions. For example, for search queries received by the electronic product catalog [i.e. a product search query by the user], the server may assign fourth tokens such as “SQ.”” [0018]),
-at least one data set (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “the server may assign second tokens to the one or more tracked actions which have catalog information or product identity within the electronic product catalog [i.e. at least one data set]. For example, activities with specific product item information, such as a selection to view an item of the electronic product catalog (viewProduct_Item), a selection to view a page of the electronic product catalog with an item (viewPage_item), and the like, may be assigned product tokens (second tokens) by the server which have associated catalog information and/or product identity” [0016]), and
-a set of experimental data (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “If there is no information available for the customer, the server may use a population mean preference vector instead. The server may align the sequence of tokens and their vectors into the model, and may generate a prediction as to what is the next most likely customer behavior [i.e. a set of experimental data]” [0046] and “Interactions by a user with an electronic product catalog may be tracked. The tracked activities may tokenized, so that there may be sequences of tokens for each user that are translated from the user's activity with an online electronic product catalog. These tokens, which may be referred to as catalog tokens, may include activity tokens [i.e. a set of experimental data], product tokens, and category tokens, and may be encoded into vectors. To make the recommendation personalized, user profile information may be encoded as a matrix of vectors” [0008]);
-determining, using at least one machine-learning algorithm (Zhang, see at least: “the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]):
-a search type (Zhang, see at least: “The implementations of the disclosed subject matter may map user interaction history and/or user behavior with the electronic product catalog to generate tokens, and the concurrencies of the tokens and/or sequence of tokens may be used to generate vectors. The interaction history and/or behavior may include product interactions, interactions with the catalog, interactions with one or more categories of items, search queries, refined search queries [i.e. a search type], and the like” [0013] and “the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]), and
-a plurality of attributes associated with the product search query (Zhang, see at least: “the server may assign third tokens to the one or more tracked actions based on at least one category of the electronic product catalog. For example, the server may assign third tokens (category tokens) for category data such as the category of current page [i.e. a plurality of attributes associated with the product search query], such as “CT1, “CT2,” . . . , “CTm.” In some implementations, a merchant may have a website with an electronic product catalog, and the third tokens (category tokens) may be based on the categories for the website of the particular merchant” [0017] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions [i.e. a plurality of attributes associated with the product search query]” [0018] and “The implementations of the disclosed subject matter may map user interaction history and/or user behavior with the electronic product catalog to generate tokens, and the concurrencies of the tokens and/or sequence of tokens may be used to generate vectors. The interaction history and/or behavior may include product interactions, interactions with the catalog, interactions with one or more categories of items, search queries [i.e. a plurality of attributes associated with the product search query], refined search queries, and the like” [0013] and “ the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]),
-using a machine learning algorithm to determine at least one top alternative product based on the plurality of attributes associated with the product search query (Zhang, see at least: “the recommended product may be generated [i.e. to determine at least one top alternative product] by the server according to a trained model of machine learning [i.e. using a machine learning algorithm to]. That is, the server may utilize: (1) the pre-learned shopping patterns of co-occurrence of different product tokens and activity tokens [i.e. based on the plurality of search attributes associated with the product search query]; and (2) the current user sequential behavior to determine one or more items to be recommended. The model may fit the current sequential behavior with the learned patterns” [0042] and “the server may assign third tokens to the one or more tracked actions based on at least one category of the electronic product catalog. For example, the server may assign third tokens (category tokens) for category data such as the category of current page [i.e. based on the plurality of search attributes associated with the product search query], such as “CT1, “CT2,” . . . , “CTm.” In some implementations, a merchant may have a website with an electronic product catalog, and the third tokens (category tokens) may be based on the categories for the website of the particular merchant” [0017] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions [i.e. based on the plurality of search attributes associated with the product search query]” [0018]);
-transmitting the at least one top alternative product for display to the user on an interactive web page of a user device (Zhang, see at least: “the recommended products 226 are shown in display 220 of FIG. 3D [i.e. transmitting the at least one top alternative product for display to the user on an interactive a web page of a user device], but the server may generate recommended products and display them on displays 200, 209, 219, 220, and/or 229 of FIGS. 3A-3E. For example, when the customer types a search query into search dialog box 204 in FIG. 3A, the server may generate recommendations and provide a list of ranked items in the search results 210 shown in FIG. 3B” [0043]); and
-transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order (Zhang, see at least: “FIGS. 3A-3E show examples of user interactions [i.e. from the user on an interface of the web page] with a product catalog, generating tokens, and providing product recommendations using the method 100 shown in FIGS. 1-2 and described above according to implementations of the disclosed subject matter … In some implementations, a user may browse, search, and/or select products for purchase in the electronic product catalog [i.e. transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order]” [0033]).
Zhang does not explicitly disclose performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features.
Sun, however, teaches searching for data (i.e. abstract), including performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features (Sun, see at least: “In order to determine a product's brand name using a finite state machine, the opinion search engine 60 may use a word breaker to generate or create a sequence of terms contained in a product name as it is listed on a web page. Given a finite state machine having states and transitions represented by terms and grammar, and an input term sequence (e.g, product) [i.e. between the product search query] that is not accepted by the finite state machine, the opinion search engine 60 may use an algorithm to find a minimal set of modifications that may be made to the finite state machine such that the input term sequence is acceptable [i.e. a set of key features]. In one implementation, the opinion search engine 60 may run the algorithm iteratively [i.e. performing an iterative match, using iterative string matching and a set of key features] update the finite state machine such that it contains the input term sequence. The opinion search engine 60 may update the finite state machine through induction such that the transition represented by the terms may be changed into transitions represented by grammar, and new transitions and new states may be added to the finite state machine if they were not previously represented on the finite state machine. The output of algorithm is a finite state machine consisting of multiple state sequences such that each state sequence is a product name. The opinion search engine 60 may then use the updated finite state machine to predict if an input term sequence corresponds with a brand name” [0026] and “the opinion search engine 60 may extract metadata from each webpage stored on the opinion database. The metadata may include information pertaining to the opinion data on the webpage stored on the opinion database [i.e. between the product search query and a product index data] … The opinion search engine 60 may determine a product's brand name by using a finite state machine” [0025]).
It would have been obvious to one of ordinary skill in the art to include in the system, as taught by Zhang, performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features, as taught by Sun, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. It further would have been obvious to one of ordinary skill in the art at the time of filing to modify Zhang, to include the teachings of Sun, in order to allow data to be easily queried (Sun, [0029]).
Zhang in view of Sun does not explicitly teach upon performing the iterative match, determining no matches between the product search query and the product index data set; based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product.
Garg, however, teaches providing content to a user based on a search query and a class of the content (i.e. abstract), including upon performing the iterative match, determining no matches between the product search query and the product index data set (Garg, see at least: “for an e-commerce site query (e.g., querying documents in the product catalog of a given merchant,50 such as an e-commerce web site for ACME Company that sells various clothing/apparel products), the following attributes can be used for matching each product [i.e. upon performing the iterative match]: product title, product description, brand of the product (e.g., clothing brand), page crumb, product category tree, color of the product, size of the product, gender of the product, and/or other attributes” Col. 11 Ln. 45-52 and “a threshold is applied based on the relevance (RVS) score calculated by the relevance subsystem to reduce the base set of results and then the results set is sent to further ranking subsystems, such as further described below. An example of such an implementation can limit the results with the top X percent of the total relevance score range to be passed onto other subsystems (e.g., applying an 80.sup.th percentile threshold or some other percentile threshold). For example, if the relevance scores of possible results for a search query ranged from 0 to 200 [i.e. determining no matches between the product search query and the product index data set], then a threshold can be applied that would only pass through the results with the top 80.sup.th percent scores (e.g., relevance scores greater than 40)” Col. 11 Ln. 53-65 Examiner notes that if the relevance score is below 200, then there isn’t an exact match; Examiner further notes that ‘no matches’ is interpreted in light of [004] of Applicant’s spec that indicates “if no exact match exists based on the product of interest, the methods or systems would automatically identify the likely product categories associated with the user's product of interest”);
the known technique of based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query (Garg, see at least: “for an e-commerce site query (e.g., querying documents in the product catalog of a given merchant, such as an e-commerce web site for ACME Company that sells various clothing/apparel products), the following attributes can be used for matching each product: product title, product description, brand of the product (e.g., clothing brand), page crumb, product category tree, color of the product, size of the product, gender of the product, and/or other attributes [i.e. based on the plurality of search attributes associated with the product search query]” Col. 11 Ln. 45-52 and “a threshold is applied based on the relevance (RVS) score calculated by the relevance subsystem to reduce the base set of results and then the results set is sent to further ranking subsystems, such as further described below. An example of such an implementation can limit the results with the top X percent of the total relevance score range to be passed onto other subsystems (e.g., applying an 80.sup.th percentile threshold or some other percentile threshold) [i.e. determine at least one top alternative product]. For example, if the relevance scores of possible results for a search query ranged from 0 to 200 [i.e. based on the determination of no matches], then a threshold can be applied that would only pass through the results with the top 80.sup.th percent scores (e.g., relevance scores greater than 40)” Col. 11 Ln. 53-65 and “This feedback-loop mechanism can provide an important signal to the query-dependent and content-class based ranking system to continuously learn [i.e. using a machine learning algorithm to] the most relevant set of results for any query apart from the content and overall (e.g., query-independent) attributes of the potential results” Col. 6 Ln. 37-42; Examiner notes that if the relevance score is below 200, then there isn’t an exact match; Examiner further notes that ‘no matches’ is interpreted in light of [004] of Applicant’s spec that indicates “if no exact match exists based on the product of interest, the methods or systems would automatically identify the likely product categories associated with the user's product of interest”); and
the known technique of the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product (Garg, see at least: “FIG. 4 illustrates an embodiment of a web page 400 that includes widgets 402 and 404 that are populated with content identified [i.e. the web page including interactive user elements regarding the top alternative product] using the disclosed techniques. In various embodiments, content (e.g., images, texts, links such as URL/URI links, and/or other static or dynamic web page content) [i.e. including at least one selectable element, a picture of the top alternative product] may be published on a web page in any appropriate manner, including in-line with previously existing content and/or separately via one or more dedicated widgets” Col. 21 Ln. 57-64 and “the query interpretation subsystem can interpret queries to determine a focus or intent of a given user, such as a user's intent to focus on a specific product, price, alternative content [i.e. regarding the top alternative product], brand, color, and/or other aspects, which can be used to determine what to return in the ranked search results as further described herein” Col. 18 Ln. 31-37 and “if a personalization module determines (e.g., deduces) based on a monitored user's activity on a merchant's web site (e.g., by using behavioral data on a merchant's web site, which can be stored in merchant web services data 230, such as discussed above), that a given user has preferences for certain brands in certain product categories, for queries that return results in these categories, then the products from those preferred brands can be boosted in the search results for that given user by applying the personalized boost factor to increase scores for such products for that given user [i.e. and an ordered list of sellers associated with the top alternative product]” Col. 12 Ln. 33-43). These known technique is applicable to the system of Zhang in view of Sun as they both share characteristics and capabilities, namely, they are directed to providing content to a user based on a search query and a class of the content.
It would have been obvious to one of ordinary skill in the art to include in the system, as taught by Zhang in view of Sun, upon performing the iterative match, determining no matches between the product search query and the product index data set, as taught by Garg, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. It further would have been obvious to one of ordinary skill in the art at the time of filing to modify Zhang in view of Sun, to include the teachings of Garg, in order to provide incremental updates to product catalog information (Garg, Col. 8 Ln. 63-67 & Col. 9 Ln. 1-2).
Additionally, it would have been recognized that applying the known techniques of, based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product, as taught by Garg, to the teachings of Zhang in view of Sun would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar systems. Further, adding the modifications of, based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product, as taught by Garg, into the system of Zhang in view of Sun would have been recognized by those of ordinary skill in the art as resulting in an improved system that would provide incremental updates to product catalog information (Garg, Col. 8 Ln. 63-67 & Col. 9 Ln. 1-2).
Regarding claim 2, the combination of Zhang/Sun/Garg teaches the system of claim 1. Zhang further discloses:
-wherein the at least one data set comprises a catalogue of product model numbers (Zhang, see at least: “The electronic product catalog may be stored on the server and/or a database system (e.g., database systems 1200 a-d shown in FIG. 5) [i.e. wherein the at least one data set comprises a catalogue] communicatively coupled to the server” [0014] and “the customer may select recommended product 228 (i.e., having catalog identification number 82) [i.e. a catalogue of product model numbers]” [0045]).
Zhang in view of Sun does not explicitly disclose that the catalogue of product model numbers are collected over a predefined time frame.
Garg further teaches providing content to a user based on a search query and a class of the content (i.e. abstract), including the known technique of at least one data set comprising a catalogue of product model numbers collected over a predefined time frame (Garg, see at least: “a process can be performed to retrieve merchant data feeds for each of the merchants (e.g., implemented using a merchant data feed retrieving engine as disclosed herein). In an example implementation, the merchant can either provide a location to fetch the merchant data feed or upload the merchant data feed to a location specified for retrieving (e.g., pulling) the merchant data feed (e.g., periodically or upon demand) [i.e. collected over a predefined time frame]. For example, the merchant data feed can be provided as a file (e.g., in a structured data format that can be parsed, such as in an XML or other file format that provides a (subset) view of, for example, a merchant's product/service catalog, such as by providing incremental updates to product catalog information using such merchant data feeds) [i.e. the at least one data set comprises a catalogue]” Col. 8 Ln. 56-67 & Col. 9 Ln. 1-2 and “different queries can be processed to determine an intent of the query, such as how a user desires to engage in a given session (e.g., shopping cycle/session). For example, a general search query can indicate an informational browsing session by a user in contrast to a more targeted search query that can indicate a targeted purchase session by a user (e.g., a search query of “television” as opposed to a search query of “42-inch LED television” or a search query that specifies a particular model number [i.e. a catalogue of product model numbers] of an LED television sold by a merchant)” Col. 17 Ln. 53-62). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Zhang in view of Sun with Garg for the reasons identified above with respect to claim 1.
Regarding claim 3, the combination of Zhang/Sun/Garg teaches the system of claim 1. Zhang further discloses:
-wherein the user product search query comprises at least an alphanumeric product model number, a text string, or a combination thereof (Zhang, see at least: “The server (e.g., central component 600, and/or second computer 700 shown in FIG. 4) may generate a search token (i.e., a fourth token as described above), which may include text content (i.e., “running shoe for male”) [i.e. wherein the user product search query comprises at least an alphanumeric product model number, a text string, or a combination thereof]. The search token may be assigned two types of vectors by the server” [0036]).
Regarding claim 5, the combination of Zhang/Sun/Garg teaches the system of claim 1. Zhang further discloses:
-wherein the data structures comprise linear data structures, or non-linear data structures (Zhang, see at least: "FIGS. 4-5 can store the data (e.g., the electronic product catalog, tracked activities, generated tokens, generated vectors, product recommendations, and the like) in the immutable storage of the at least one storage device (e.g., a storage device associated with central component 600, the second computer 700, and/or the database systems 1200 a-1200 d) using a log-structured merge tree data structure [i.e. wherein the data structures comprise linear data structures, or non-linear data structure]" [0056]).
Regarding claim 6, the combination of Zhang/Sun/Garg teaches the system of claim 1. Zhang further discloses:
-wherein the plurality of attributes associated with the product query comprises a product model number, a product name, or product description (Zhang, see at least: “The server (e.g., central component 600, and/or second computer 700 shown in FIG. 4) may generate a search token (i.e., a fourth token as described above), which may include text content (i.e., “running shoe for male [i.e. wherein the plurality of attributes associated with the product query comprises a product model number, a product name, or product description]”)” [0036]).
Claims 11-12 and 14-15 recite limitations directed towards a computer-implemented method. The limitations recited in claims 11-12 and 14-15 are parallel in nature to those addressed above for claims 1-2 and 5-6, respectively, and are therefore rejected for those same reasons set forth above in claims 1-2 and 5-6, respectively.
Regarding claim 17, the combination of Zhang/Sun/Garg teaches the method of claim 11. Zhang further discloses:
-wherein the determination of the at least one top alternative product is based on a queried product category (Zhang, see at least: “the recommended product may be generated [i.e. determining the at least one top alternative product] by the server according to a trained model of machine learning. That is, the server may utilize: (1) the pre-learned shopping patterns of co-occurrence of different product tokens and activity tokens; and (2) the current user sequential behavior to determine one or more items to be recommended. The model may fit the current sequential behavior with the learned patterns [i.e. based on a queried product category]” [0042] and “the server may assign third tokens to the one or more tracked actions based on at least one category of the electronic product catalog. For example, the server may assign third tokens (category tokens) for category data such as the category of current page [i.e. based on the queried product category], such as “CT1, “CT2,” . . . , “CTm.” In some implementations, a merchant may have a website with an electronic product catalog, and the third tokens (category tokens) may be based on the categories for the website of the particular merchant” [0017] and “the server may generate a language model by determining the co-occurrence of one or more of the first tokens, the second tokens, the third tokens, and/or the fourth tokens. At operation 184, the server may generate the at least one recommendation based on the determined co-occurrence of tokens [i.e. based on a queried product category] using the generated language model” [0022]).
Claims 4, 7-9, 13, 16, 18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al. (US 2021/0358007 A1), hereinafter Zhang, in view of Sun, in further view of Garg, in further view of Tavernier et al. (US 9,904,949 A1), hereinafter Tavernier.
Regarding claim 4, the combination of Zhang/Sun/Garg teaches the system of claim 1. Zhang further discloses:
The combination of Zhang/Sun/Garg does not explicitly teach the experimental data comprising at least aggregated purchase data from all customers or a subset of all customers.
Tavernier, however, teaches recommending similar products (i.e. abstract), including the known technique of the experimental data comprising at least aggregated purchase data from all customers or a subset of all customers (Tavernier, see at least: “customer segments may be used as a basis for identifying or selecting similarities datasets for recommendations. As an example, customers that visit the electronic retail store frequently and spend over a threshold monetary value may have a preference for session similarities over purchase similarities or vice versa [i.e. wherein the experimental data comprises at least aggregated purchase data from all customers or a subset of all customers]” Col. 5 Ln. 38-44). This known technique is applicable to the system of the combination of Zhang/Sun/Garg as they both share characteristics and capabilities, namely, they are directed to recommending similar products.
It would have been recognized that applying the known technique of the experimental data comprising at least aggregated purchase data from all customers or a subset of all customers, as taught by Tavernier, to the teachings of the combination of Zhang/Sun/Garg would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar systems. Further, adding the modification of the experimental data comprising at least aggregated purchase data from all customers or a subset of all customers, as taught by Tavernier, into the system of the combination of Zhang/Sun/Garg would have been recognized by those of ordinary skill in the art as resulting in an improved system that would result in increased sales, improved brand awareness, and greater transaction satisfaction by the customer (Tavernier, Col. 1 Ln. 26-30).
Regarding claim 7, the combination of Zhang/Sun/Garg teaches the system of claim 1.
Zhang in view of Sun does not explicitly teach the operations further include determining the set of key features based on the experimental data, wherein the determination of the key features associated with the queried product is further based on mined data from at least one external data source.
Garg further teaches providing content to a user based on a search query and a class of the content (i.e. abstract), including the known technique of the determination of the key features associated with the queried product being based on mined data from at least one external data source (Garg, see at least: “the above-described data sources (e.g., data sources 110 such as for a given merchant and/or other data sources for other merchants) [i.e. from at least one external data source] can be processed using various techniques as described below. For example, a web site content fetching process can be performed (e.g., implemented using a web site content fetching engine) to fetch web pages from a merchant web site (e.g., to periodically (crawl and) [i.e. is further based on mined data] fetch web pages from merchant web site 106, and also shown as data source, merchant web site 112 of merchant data sources 110). The content (e.g., HTML content, dynamic content such as Java Script content, and/or other content) from the fetched web pages from the merchant web site can then be parsed to extract attributes for a product, such as a title, a heading, a description, a price, a color(s), and/or a size(s). As another example, a merchant data feed retrieving process can be performed (e.g., implemented using a merchant data feed retrieving engine) to retrieve merchant content/product data feeds (e.g., merchant data feed 114) and parse product attributes, such as title, heading, description, and/or price. In some cases, a merchant data feed can include a content classification (e.g., a merchant product data feed can include a product classification) [i.e. the determination of the key features associated with the queried product]” Col. 7 Ln. 64-67 & Col. 8 Ln. 1-18). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Zhang in view of Sun with Garg for the reasons identified above with respect to claim 1.
The combination of Zhang/Sun/Garg does not explicitly teach determining the set of key features based on the experimental data.
Tavernier, however, teaches recommending similar products (i.e. abstract), including the known technique of determining the set of key features based on the experimental data (Tavernier, see at least: “The model may be used to pick a desired similarities dataset to use given a context, as identified from context items, such as input source signal, customer or user identification, page context and widget. Other context items may include, for example, a date, a time, a geographic location of a user, and a product identification, price, brand or category [i.e. determining the set of key features]” Col. 5 Ln. 4-10 and “customer segments may be used as a basis for identifying or selecting similarities datasets for recommendations. As an example, customers that visit the electronic retail store frequently and spend over a threshold monetary value may have a preference for session similarities over purchase similarities or vice versa [i.e. based on the experimental data]” Col. 5 Ln. 38-44 and “use of certain recommendations may be more beneficial for certain categories of products [i.e. determining the set of key features] than recommendations from a different source or based on a different type of similarity” Col. 4 Ln. 37-40 and “If the user has recently purchased the electronic device, then similarly, accessories, software and the like which are compatible with the electronic device may be appropriate recommendations [i.e. determining the set of key features]. If the user is looking to purchase an electronic device, a purchase similarities dataset including accessories may be a source of recommendations. However, in some examples, displaying accessories for a not-yet-purchased device may be premature. Machine learned models may thus be used to determine which similarities dataset to use as the basis for recommendations based on page context [i.e. based on the experimental data]” Col. 6 Ln. 62-67 & Col. 7 Ln. 1-5). This known technique is applicable to the system of the combination of Zhang/Sun/Garg as they both share characteristics and capabilities, namely, they are directed to recommending similar products.
It would have been recognized that applying the known technique of determining the set of key features based on the experimental data, as taught by Tavernier, to the teachings of the combination of Zhang/Sun/Garg would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar systems. Further, adding the modification of determining the set of key features based on the experimental data, as taught by Tavernier, into the system of the combination of Zhang/Sun/Garg would have been recognized by those of ordinary skill in the art as resulting in an improved system that would result in increased sales, improved brand awareness, and greater transaction satisfaction by the customer (Tavernier, Col. 1 Ln. 26-30).
Regarding claim 8, the combination of Zhang/Sun/Garg teaches the system of claim 1.
The combination of Zhang/Sun/Garg does not explicitly teach the determination of the at least one top alternative product being based on a queried product category and an associated set of pre-determined rules.
Tavernier further teaches recommending similar products (i.e. abstract), including the known technique of the determination of the at least one top alternative product being based on a queried product category and an associated set of pre-determined rules (Tavernier, see at least: “The model may be used to pick a desired similarities dataset to use given a context, as identified from context items, such as input source signal, customer or user identification, page context and widget [i.e. the determination of the at least one top alternative product is based on an associated set of pre-determined rules]. Other context items may include, for example, a date, a time, a geographic location of a user, and a product identification, price, brand or category [i.e. is based on the queried product category]” Col. 5 Ln. 4-10 and “customer segments may be used as a basis for identifying or selecting similarities datasets for recommendations [i.e. wherein the determination of the at least one top alternative product]. As an example, customers that visit the electronic retail store frequently and spend over a threshold monetary value may have a preference for session similarities over purchase similarities or vice versa” Col. 5 Ln. 38-44 and “use of certain recommendations may be more beneficial for certain categories of products than recommendations from a different source or based on a different type of similarity [i.e. is based on the queried product category and an associated set of pre-determined rules]” Col. 4 Ln. 37-40 and “the selection model 270 may select purchase similarities as a basis of recommendations unless there is a particular rule prohibiting the selection or unless purchase similarity-based recommendations are already displayed on the content page. In the event of an exception, another similarities dataset may be used, such as the session similarities dataset. Likewise, rules [i.e. an associated set of pre-determined rules] may be provided which may allow or prohibit the use of the session similarities dataset” Col. 7 Ln. 35-43). This known technique is applicable to the system of the combination of Zhang/Sun/Garg as they both share characteristics and capabilities, namely, they are directed to recommending similar products.
It would have been recognized that applying the known technique of the determination of the at least one top alternative product being based on a queried product category and an associated set of pre-determined rules, as taught by Tavernier, to the teachings of the combination of Zhang/Sun/Garg would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar systems. Further, adding the modification of the determination of the at least one top alternative product being based on a queried product category and an associated set of pre-determined rules, as taught by Tavernier, into the system of the combination of Zhang/Sun/Garg would have been recognized by those of ordinary skill in the art as resulting in an improved system that would result in increased sales, improved brand awareness, and greater transaction satisfaction by the customer (Tavernier, Col. 1 Ln. 26-30).
Regarding claim 9, the combination of Zhang/Sun/Garg/Tavernier teaches the system of claim 7.
The combination of Zhang/Sun/Garg does not explicitly disclose the determination of the at least one top alternative product is based on an inference relating to the set of key features associated with the product.
Tavernier further teaches recommending similar products (i.e. abstract), including the known technique of the determination of the at least one top alternative product being based on an inference relating to the set of key features associated with the product (Tavernier, see at least: “use of certain recommendations may be more beneficial for certain categories of products than recommendations from a different source or based on a different type of similarity [i.e. the determination of the at least one top alternative product is based on an inference relating to the set of key features associated with the product]” Col. 4 Ln. 37-40 and “If the user has recently purchased the electronic device, then similarly, accessories, software and the like which are compatible with the electronic device may be appropriate recommendations [i.e. the determination of the at least one top alternative product is based on an inference relating to the set of key features associated with the product]. If the user is looking to purchase an electronic device, a purchase similarities dataset including accessories may be a source of recommendations. However, in some examples, displaying accessories for a not-yet-purchased device may be premature. Machine learned models may thus be used to determine which similarities dataset to use as the basis for recommendations based on page context” Col. 6 Ln. 62-67 & Col. 7 Ln. 1-5 Examiner notes that the key features include a type of a software system for an electronic device and the inference is the compatibility of the accessories). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Zhang in view of Garg with Tavernier for the reasons identified above with respect to claim 7.
Claims 13 and 16 recite limitations directed towards a computer-implemented method. The limitations recited in claims 13 and 16 are parallel in nature to those addressed above for claims 4 and 7 respectively, and are therefore rejected for those same reasons set forth above in claims 4 and 7, respectively.
Regarding claim 18, the combination of Zhang/Sun/Garg/Tavernier teaches the method of claim 16.
The combination of Zhang/Sun/Garg/Tavernier does not explicitly disclose the determination of the at least one top alternative product being based on the set of key features associated with the product.
Tavernier further teaches recommending similar products (i.e. abstract), including the known technique of the determination of the at least one top alternative product being based on the set of key features associated with the product (Tavernier, see at least: “use of certain recommendations may be more beneficial for certain categories of products than recommendations from a different source or based on a different type of similarity [i.e. the determination of the at least one top alternative product is based on the set of key features associated with the product]” Col. 4 Ln. 37-40 and “If the user has recently purchased the electronic device, then similarly, accessories, software and the like which are compatible with the electronic device may be appropriate recommendations [i.e. the determination of the at least one top alternative product is based on the set of key features associated with the product]. If the user is looking to purchase an electronic device, a purchase similarities dataset including accessories may be a source of recommendations. However, in some examples, displaying accessories for a not-yet-purchased device may be premature. Machine learned models may thus be used to determine which similarities dataset to use as the basis for recommendations based on page context” Col. 6 Ln. 62-67 & Col. 7 Ln. 1-5 Examiner notes that the key features include a type of a software system for an electronic device). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the combination of Zhang/Sun/Garg/Tavernier with Tavernier for the reasons identified above with respect to claim 16.
Regarding claim 20, Zhang discloses a computer-implemented system for identifying a target product and generating alternative product recommendations based on a user query, the system comprising:
-a memory storing instructions (Zhang, see at least: “The storage 710 may use any suitable combination of any suitable volatile and non-volatile physical storage mediums, including, for example, hard disk drives, solid state drives, optical media, flash memory, tape drives, registers, and random access memory, or the like, or any combination thereof” [0047]); and
-at least one processor configured to execute the instructions to perform operations (Zhang, see at least: “When implemented on a general-purpose microprocessor, the computer program code segments configure the microprocessor to create specific logic circuits” [0061]) comprising:
-retrieving, from one or more data structures (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048]):
-a product search query by a user comprising at least an alphanumeric product model number, a text string, or any combination thereof (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions. For example, for search queries received by the electronic product catalog [i.e. a product search query by the user], the server may assign fourth tokens such as “SQ.”” [0018] and “The server (e.g., central component 600, and/or second computer 700 shown in FIG. 4) may generate a search token (i.e., a fourth token as described above), which may include text content (i.e., “running shoe for male”) [i.e. comprising at least an alphanumeric product model number, a text string, or any combination thereof]. The search token may be assigned two types of vectors by the server” [0036]),
-at least one data set comprising at least a catalogue of product model numbers (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “The electronic product catalog may be stored on the server and/or a database system (e.g., database systems 1200 a-d shown in FIG. 5) [i.e. at least one data set comprising at least a catalogue] communicatively coupled to the server” [0014] and “the customer may select recommended product 228 (i.e., having catalog identification number 82) [i.e. a catalogue of product model numbers]” [0045]), and
-a set of experimental data (Zhang, see at least: “The storage 710 of the second computer 700 can store data (e.g., data that is part of the electronic product catalog, generated tokens, generated vectors, tracked activities, or the like) [i.e. retrieving, from one or more data structures]” [0048] and “If there is no information available for the customer, the server may use a population mean preference vector instead. The server may align the sequence of tokens and their vectors into the model, and may generate a prediction as to what is the next most likely customer behavior [i.e. a set of experimental data]” [0046] and “Interactions by a user with an electronic product catalog may be tracked. The tracked activities may tokenized, so that there may be sequences of tokens for each user that are translated from the user's activity with an online electronic product catalog. These tokens, which may be referred to as catalog tokens, may include activity tokens [i.e. a set of experimental data], product tokens, and category tokens, and may be encoded into vectors. To make the recommendation personalized, user profile information may be encoded as a matrix of vectors” [0008]);
-determining, using at least one machine-learning algorithm (Zhang, see at least: “the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]):
-a search type (Zhang, see at least: “The implementations of the disclosed subject matter may map user interaction history and/or user behavior with the electronic product catalog to generate tokens, and the concurrencies of the tokens and/or sequence of tokens may be used to generate vectors. The interaction history and/or behavior may include product interactions, interactions with the catalog, interactions with one or more categories of items, search queries, refined search queries [i.e. a search type], and the like” [0013] and “the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]), and
-a plurality of attributes associated with the product query comprising at least a product model number, a product name, or product description (Zhang, see at least: “the server may assign third tokens to the one or more tracked actions based on at least one category of the electronic product catalog. For example, the server may assign third tokens (category tokens) for category data such as the category of current page [i.e. a plurality of attributes associated with the product search query], such as “CT1, “CT2,” . . . , “CTm.” In some implementations, a merchant may have a website with an electronic product catalog, and the third tokens (category tokens) may be based on the categories for the website of the particular merchant” [0017] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions [i.e. a plurality of attributes associated with the product search query]” [0018] and “The server (e.g., central component 600, and/or second computer 700 shown in FIG. 4) may generate a search token (i.e., a fourth token as described above), which may include text content (i.e., “running shoe for male [i.e. comprising at least a product model number, a product name, or product description]”)” [0036] and “The implementations of the disclosed subject matter may map user interaction history and/or user behavior with the electronic product catalog to generate tokens, and the concurrencies of the tokens and/or sequence of tokens may be used to generate vectors. The interaction history and/or behavior may include product interactions, interactions with the catalog, interactions with one or more categories of items, search queries [i.e. a plurality of attributes associated with the product search query], refined search queries, and the like” [0013] and “the language model may be generated by the server generating two sub-models for each set of for the second tokens (product tokens) and the third tokens (category tokens). The server may generate a densely connected two-layer sub-model and a softmax outer layer (i.e., final layer). The sub-models generated for the second tokens may be generated by the server applying an embedding layering for vectors formed from the second tokens” [0024] and “the sub-models may be bidirectional long short-term memory (LSTM) neural networks with attention (e.g., where dependencies between source and target sequences are not restricted by the in-between distance), which may include feedback connections [i.e. determining, using at least one machine-learning algorithm:]” [0025]),
-using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query (Zhang, see at least: “the recommended product may be generated [i.e. to determine at least one top alternative product] by the server according to a trained model of machine learning [i.e. using a machine learning algorithm to]. That is, the server may utilize: (1) the pre-learned shopping patterns of co-occurrence of different product tokens and activity tokens [i.e. based on the plurality of search attributes associated with the product search query]; and (2) the current user sequential behavior to determine one or more items to be recommended. The model may fit the current sequential behavior with the learned patterns” [0042] and “the server may assign third tokens to the one or more tracked actions based on at least one category of the electronic product catalog. For example, the server may assign third tokens (category tokens) for category data such as the category of current page [i.e. based on the plurality of search attributes associated with the product search query], such as “CT1, “CT2,” . . . , “CTm.” In some implementations, a merchant may have a website with an electronic product catalog, and the third tokens (category tokens) may be based on the categories for the website of the particular merchant” [0017] and “the server may assign fourth tokens to search queries and/or search refinement of the one or more tracked actions [i.e. based on the plurality of search attributes associated with the product search query]” [0018]);
-transmitting the at least one top alternative product for display to the user on an interactive web page of a user device (Zhang, see at least: “the recommended products 226 are shown in display 220 of FIG. 3D [i.e. transmitting the at least one top alternative product for display to the user on an interactive web page of a user device], but the server may generate recommended products and display them on displays 200, 209, 219, 220, and/or 229 of FIGS. 3A-3E. For example, when the customer types a search query into search dialog box 204 in FIG. 3A, the server may generate recommendations and provide a list of ranked items in the search results 210 shown in FIG. 3B” [0043]); and
-transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order (Zhang, see at least: “FIGS. 3A-3E show examples of user interactions [i.e. from the user on an interface of the web page] with a product catalog, generating tokens, and providing product recommendations using the method 100 shown in FIGS. 1-2 and described above according to implementations of the disclosed subject matter … In some implementations, a user may browse, search, and/or select products for purchase in the electronic product catalog [i.e. transmitting a selection from the user on an interface of the web page, to the system, wherein the selection includes a new order]” [0033]).
Zhang does not explicitly disclose performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features.
Sun, however, teaches searching for data (i.e. abstract), including performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features (Sun, see at least: “In order to determine a product's brand name using a finite state machine, the opinion search engine 60 may use a word breaker to generate or create a sequence of terms contained in a product name as it is listed on a web page. Given a finite state machine having states and transitions represented by terms and grammar, and an input term sequence (e.g, product) [i.e. between the product search query] that is not accepted by the finite state machine, the opinion search engine 60 may use an algorithm to find a minimal set of modifications that may be made to the finite state machine such that the input term sequence is acceptable [i.e. a set of key features]. In one implementation, the opinion search engine 60 may run the algorithm iteratively [i.e. performing an iterative match, using iterative string matching and a set of key features] update the finite state machine such that it contains the input term sequence. The opinion search engine 60 may update the finite state machine through induction such that the transition represented by the terms may be changed into transitions represented by grammar, and new transitions and new states may be added to the finite state machine if they were not previously represented on the finite state machine. The output of algorithm is a finite state machine consisting of multiple state sequences such that each state sequence is a product name. The opinion search engine 60 may then use the updated finite state machine to predict if an input term sequence corresponds with a brand name” [0026] and “the opinion search engine 60 may extract metadata from each webpage stored on the opinion database. The metadata may include information pertaining to the opinion data on the webpage stored on the opinion database [i.e. between the product search query and a product index data] … The opinion search engine 60 may determine a product's brand name by using a finite state machine” [0025]).
It would have been obvious to one of ordinary skill in the art to include in the method, as taught by Zhang, performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features, as taught by Sun, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. It further would have been obvious to one of ordinary skill in the art at the time of filing to modify Zhang, to include the teachings of Sun, in order to allow data to be easily queried (Sun, [0029]).
Zhang in view of Sun does not explicitly teach Zhang does not explicitly disclose the at least one data set comprising a catalogue of product model numbers collected over a predefined time frame; upon performing the iterative match, determining no matches between the product search query and the product index data set; based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product.
Garg, however, teaches providing content to a user based on a search query and a class of the content (i.e. abstract), including the known technique of at least one data set comprising a catalogue of product model numbers collected over a predefined time frame (Garg, see at least: “a process can be performed to retrieve merchant data feeds for each of the merchants (e.g., implemented using a merchant data feed retrieving engine as disclosed herein). In an example implementation, the merchant can either provide a location to fetch the merchant data feed or upload the merchant data feed to a location specified for retrieving (e.g., pulling) the merchant data feed (e.g., periodically or upon demand) [i.e. collected over a predefined time frame]. For example, the merchant data feed can be provided as a file (e.g., in a structured data format that can be parsed, such as in an XML or other file format that provides a (subset) view of, for example, a merchant's product/service catalog, such as by providing incremental updates to product catalog information using such merchant data feeds) [i.e. the at least one data set comprises a catalogue]” Col. 8 Ln. 56-67 & Col. 9 Ln. 1-2 and “different queries can be processed to determine an intent of the query, such as how a user desires to engage in a given session (e.g., shopping cycle/session). For example, a general search query can indicate an informational browsing session by a user in contrast to a more targeted search query that can indicate a targeted purchase session by a user (e.g., a search query of “television” as opposed to a search query of “42-inch LED television” or a search query that specifies a particular model number [i.e. a catalogue of product model numbers] of an LED television sold by a merchant)” Col. 17 Ln. 53-62);
upon performing the iterative match, determining no matches between the product search query and the product index data set (Garg, see at least: “for an e-commerce site query (e.g., querying documents in the product catalog of a given merchant, such as an e-commerce web site for ACME Company that sells various clothing/apparel products), the following attributes can be used for matching each product [i.e. upon performing the iterative match]: product title, product description, brand of the product (e.g., clothing brand), page crumb, product category tree, color of the product, size of the product, gender of the product, and/or other attributes” Col. 11 Ln. 45-52 and “a threshold is applied based on the relevance (RVS) score calculated by the relevance subsystem to reduce the base set of results and then the results set is sent to further ranking subsystems, such as further described below. An example of such an implementation can limit the results with the top X percent of the total relevance score range to be passed onto other subsystems (e.g., applying an 80.sup.th percentile threshold or some other percentile threshold). For example, if the relevance scores of possible results for a search query ranged from 0 to 200 [i.e. determining no matches between the product search query and the product index data set], then a threshold can be applied that would only pass through the results with the top 80.sup.th percent scores (e.g., relevance scores greater than 40)” Col. 11 Ln. 53-65 Examiner notes that if the relevance score is below 200, then there isn’t an exact match; Examiner further notes that ‘no matches’ is interpreted in light of [004] of Applicant’s spec that indicates “if no exact match exists based on the product of interest, the methods or systems would automatically identify the likely product categories associated with the user's product of interest”);
the known technique of based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query (Garg, see at least: “for an e-commerce site query (e.g., querying documents in the product catalog of a given merchant, such as an e-commerce web site for ACME Company that sells various clothing/apparel products), the following attributes can be used for matching each product: product title, product description, brand of the product (e.g., clothing brand), page crumb, product category tree, color of the product, size of the product, gender of the product, and/or other attributes [i.e. based on the plurality of search attributes associated with the product search query]” Col. 11 Ln. 45-52 and “a threshold is applied based on the relevance (RVS) score calculated by the relevance subsystem to reduce the base set of results and then the results set is sent to further ranking subsystems, such as further described below. An example of such an implementation can limit the results with the top X percent of the total relevance score range to be passed onto other subsystems (e.g., applying an 80.sup.th percentile threshold or some other percentile threshold) [i.e. determine at least one top alternative product]. For example, if the relevance scores of possible results for a search query ranged from 0 to 200 [i.e. based on the determination of no matches], then a threshold can be applied that would only pass through the results with the top 80.sup.th percent scores (e.g., relevance scores greater than 40)” Col. 11 Ln. 53-65 and “This feedback-loop mechanism can provide an important signal to the query-dependent and content-class based ranking system to continuously learn [i.e. using a machine learning algorithm to] the most relevant set of results for any query apart from the content and overall (e.g., query-independent) attributes of the potential results” Col. 6 Ln. 37-42; Examiner notes that if the relevance score is below 200, then there isn’t an exact match; Examiner further notes that ‘no matches’ is interpreted in light of [004] of Applicant’s spec that indicates “if no exact match exists based on the product of interest, the methods or systems would automatically identify the likely product categories associated with the user's product of interest”); and
the known technique of the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product (Garg, see at least: “FIG. 4 illustrates an embodiment of a web page 400 that includes widgets 402 and 404 that are populated with content identified [i.e. the web page including interactive user elements regarding the top alternative product] using the disclosed techniques. In various embodiments, content (e.g., images, texts, links such as URL/URI links, and/or other static or dynamic web page content) [i.e. including at least one selectable element, a picture of the top alternative product] may be published on a web page in any appropriate manner, including in-line with previously existing content and/or separately via one or more dedicated widgets” Col. 21 Ln. 57-64 and “the query interpretation subsystem can interpret queries to determine a focus or intent of a given user, such as a user's intent to focus on a specific product, price, alternative content [i.e. regarding the top alternative product], brand, color, and/or other aspects, which can be used to determine what to return in the ranked search results as further described herein” Col. 18 Ln. 31-37 and “if a personalization module determines (e.g., deduces) based on a monitored user's activity on a merchant's web site (e.g., by using behavioral data on a merchant's web site, which can be stored in merchant web services data 230, such as discussed above), that a given user has preferences for certain brands in certain product categories, for queries that return results in these categories, then the products from those preferred brands can be boosted in the search results for that given user by applying the personalized boost factor to increase scores for such products for that given user [i.e. and an ordered list of sellers associated with the top alternative product]” Col. 12 Ln. 33-43).These known techniques are applicable to the system of Zhang as they both share characteristics and capabilities, namely, they are directed to providing content to a user based on a search query and a class of the content.
It would have been obvious to one of ordinary skill in the art to include in the system, as taught by Zhang, upon performing the iterative match, determining no matches between the product search query and the product index data set, as taught by Garg, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. It further would have been obvious to one of ordinary skill in the art at the time of filing to modify Zhang, to include the teachings of Garg, in order to provide incremental updates to product catalog information (Garg, Col. 8 Ln. 63-67 & Col. 9 Ln. 1-2).
Additionally, it would have been recognized that applying the known techniques of at least one data set comprising a catalogue of product model numbers collected over a predefined time frame; based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product, as taught by Garg, to the teachings of Zhang would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar systems. Further, adding the modifications of at least one data set comprising a catalogue of product model numbers collected over a predefined time frame; based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query; and the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product, as taught by Garg, into the system of Zhang would have been recognized by those of ordinary skill in the art as resulting in an improved system that would provide incremental updates to product catalog information (Garg, Col. 8 Ln. 63-67 & Col. 9 Ln. 1-2).
The combination of Zhang/Sun/Garg does not explicitly disclose the set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers.
Tavernier, however, teaches recommending similar products (i.e. abstract), including the known technique of a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers (Tavernier, see at least: “customer segments may be used as a basis for identifying or selecting similarities datasets for recommendations. As an example, customers that visit the electronic retail store frequently and spend over a threshold monetary value may have a preference for session similarities over purchase similarities or vice versa [i.e. a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers]” Col. 5 Ln. 38-44). This known technique is applicable to the system of the combination of Zhang/Sun/Garg as they both share characteristics and capabilities, namely, they are directed to recommending similar products.
It would have been recognized that applying the known technique of a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers, as taught by Tavernier, to the teachings of the combination of Zhang/Sun/Garg would have yielded predictable results because the level of ordinary skill in the art demonstrated by the references applied shows the ability to incorporate such references into similar system. Further, adding the modification of a set of experimental data comprising at least aggregated customer data from all customers or a subset of all customers, as taught by Tavernier, into the system of the combination of Zhang/Sun/Garg would have been recognized by those of ordinary skill in the art as resulting in an improved system that would result in increased sales, improved brand awareness, and greater transaction satisfaction by the customer (Tavernier, Col. 1 Ln. 26-30).
Subject Matter Allowable Over the Art
Dependent Claims 10 and 19
The following is an examiner’s statement of reasons for allowance over 35 U.S.C. §103:
In the present application, dependent claims 10 and 19 would be allowable if rewritten or amended to overcome the 35 U.S.C. §101 rejection set forth in this Office action. The following is the Examiner's statement of reasons of allowance over the prior art:
Regarding 35 U.S.C. §103, upon review of the evidence at hand, it is hereby concluded that the totality of the evidence, alone or in combination, neither anticipates, reasonably teaches, nor renders obvious the below noted features of the applicant’s invention.
Dependent claim 10 and 19 is allowable over 35 USC § 103 as follows:
The most relevant prior art made of record includes previously cited Zhang et al. (US 2021/0358007 A1), newly cited Sun et al. (US 2011/0078157 A1), previously cited Garg et al. (US 9,830,392 B1), previously cited Tavernier et al. (US 9,904,949 A1), and previously cited Sullivan et al. (US 11,361,331 B2).
The combination of Zhang/Sun/Garg teaches the system of claim 1 and the method of claim 19.
As written, dependent claims 10 and 19 require the determination of the at least one top alternative product is based on key features and product category of a second product which has the highest search frequency by customers immediately prior to the search of the queried product.
Regarding Tavernier, Tavernier teaches recommending similar products (i.e. abstract), including the known technique of the determination of at least one top alternative product being based on key features and product category of a second product (Tavernier, see at least: Col. 5 Ln. 4-10 and Col. 6 Ln. 62-67 & Col. 7 Ln. 1-5).
Though disclosing these features, Tavernier does not disclose or render obvious the features discussed above.
Regarding Sullivan, Sullivan teaches a second product which has the highest search frequency by customers immediately prior to the search of the queried product (Sullivan, see at least: Col. 2 Ln. 28-45).
Though disclosing these features, Sullivan does not disclose or render obvious the features discussed above.
Ultimately, the particular combination of limitations as claimed in dependent claims 10 and 19, is not anticipated nor rendered obvious in view of Zhang, Sun, Garg, Tavernier, and Sullivan, and the totality of the prior art. While certain references may disclose more general concepts and parts of the claim, the prior art available does not specifically disclose the particular combination of these limitations.
Zhang, Sun, Garg, Tavernier, and Sullivan, however, do not teach or suggest, alone or in combination the claimed invention. Examiner emphasizes that the prior art/additional art would only be combined and deemed obvious based on knowledge gleaned from the applicant’s disclosure. Such a reconstruction is improper (i.e. hindsight reasoning). See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971).
Cited NPL Doring (reference U cited 01/29/2026 in PTO-892) teaches optimizing a catalog search, but does not teach or suggest the recited claims.
The Examiner further emphasizes the claims as a whole and hereby asserts that the totality of the evidence fails to set forth, either explicitly or implicitly, an appropriate rationale for further modification of the evidence at hand to arrive at the claimed invention. The combination of features as claimed would not be obvious to one of ordinary skill in the art as combining various references from the totality of evidence to reach the combination of features as claimed would be a substantial reconstruction of Applicant’s claimed invention relying on improper hindsight bias.
It is thereby asserted by Examiner that, in light of the above and further deliberation over all of the evidence at hand, that the claims are allowable as the evidence at hand does not anticipate the claims and does not render obvious any further modification of the references to a person of ordinary skill in the art.
Response to Arguments
Rejections under 35 U.S.C. §101
Applicant argues that the rejection under 35 U.S.C. § 101 should be withdrawn because Applicant's claims are not directed to the abstract idea of "certain methods of organizing human activity." Office Action at 4. Applicant respectfully traverses the rejection, for at least the reason that the amended claims are "not directed to an abstract idea, but rather an improvement in computer functionality" and that improvements in computer functionality do no not constitute an abstract idea (see MPEP 2106.05(a)(I)). The pending claims, when properly construed, recite elements that are not merely methods of organizing human activity. In the Response to Arguments, the Office states that the claims recite "commercial interactions in that they encompass advertising, and marketing or sales activities." Office Action at 52-53. Applicant respectfully disagrees with this characterization of the claims. While the specification and figures may illustrate an e-commerce platform, this is not indicative that the recited claim elements are limited to "sales activities," as the Office suggests. In the MPEP and Subject Matter Eligibility Guidance ("the 2019 Update"), the Office distinguishes between claims that involve abstract ideas from claims that recite abstract ideas. In the 2019 Update, the Office emphasized that "a claim recites a judicial exception when the judicial exception is 'set forth' or 'described' in the claim." 2019 Update at 1 (Remarks, pages 11-12).
Examiner respectfully disagrees. Under Prong 1 of Step 2A in MPEP 2106.04, it is determined whether the claim recites a judicial exception. The amended claims recite the concept of determining and providing an alternative product for purchase, which falls within the “Certain Methods of Organizing Human Activity” groupings of abstract ideas, enumerated in MPEP 2106.04(a), as the recited abstract idea covers a commercial interaction in that it encompasses advertising, and marketing or sales activities. The claims themselves recite the abstract idea of determining and providing an alternative product for purchase, as indicated by the bolded portions of the claim under prong 1 in the rejection above. Accordingly, the claims are directed to an abstract idea.
Applicant argues that Applicant's claims are not directed to an "abstract idea," but rather are rooted in a technical solution for identifying specific products for a user in response to a determination that no matches were found for a user search query. For example, the claim elements recite an iteration of web page interface interactions that explain the back end process for identifying specific products for a user. At least the described additional claim elements are not directed to nor limited to sales activities or behaviors. Instead, the claimed subject matter is rooted in improvements to the user interface by determining that no matches were found and figuring out a determination of "at least one top alternative product," using the machine learning algorithm and then presenting those products via the interface (Remarks, pages 11-12).
Examiner respectfully disagrees. Under Prong 1 of Step 2A in MPEP 2106.04, it is determined whether the claim recites a judicial exception. The amended claims recite the concept of determining and providing an alternative product for purchase, which falls within the “Certain Methods of Organizing Human Activity” groupings of abstract ideas, enumerated in MPEP 2106.04(a), as the recited abstract idea covers a commercial interaction in that it encompasses advertising, and marketing or sales activities. It Under Prong 2 of Step 2A, the additional elements of “using a machine learning algorithm” “an interactive web page” “a user device,” and “interactive user elements” recited in the argued limitations are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than mere instructions to implement or apply the abstract idea on a generic computing hardware (or, merely use a computer as a tool to perform an abstract idea) and do no more than generally link the use of the judicial exception to a particular technological environment or field of use. Additionally, determining that no matches were found and figuring out a determination of "at least one top alternative product" do not improve the interface itself and merely utilizing the machine learning algorithm and the interface amount to nothing more than mere instructions to implement or apply the abstract idea on a generic computing hardware (or, merely use a computer as a tool to perform an abstract idea). Accordingly, the claims are directed to an abstract idea.
Applicant further argues that amended independent claims 10 and 20, although different in scope from claim 1, recite features similar to those discussed above with respect to claim 1 and are also eligible because they do not recite a judicial exception (Remarks, page 13).
Examiner respectfully disagrees. As detailed in response to the arguments above, claim 1 is ineligible. Accordingly, claims 10 and 20 are ineligible.
Applicant further argues that, under Step 2A, Prong 2, assuming, arguendo, that Applicant's amended claims recite an abstract idea, which Applicant does not concede, the amended claims integrate the purported judicial exception into a practical application because they recite elements that provide a specific improvement over prior systems. Claim 1 integrates the alleged abstract idea into a practical application. Using conventional techniques, customizing a product recommendation based on a user's product of interesting and identified key features was not possible. If the original product of interest was not identified using conventional techniques, then those techniques were not capable of generating alternative product recommendations. See Specification paragraphs [003]-[004]. In the Response to Arguments, the Office states that customizing product recommendations "is not a technical field." Office Action at 54. Applicant respectfully disagrees with this characterization of the claims. The claims are rooted in computer technology that improves existing techniques for customizing alternative product recommendations (Remarks, pages 13-14).
Examiner respectfully disagrees. Again, customizing product recommendations is not a technical field (see MPEP 2106.05(a)). Merely reciting generic computer technology to improve the abstract idea (not the technology itself) fails to reflect an improvement in the functioning of a computer or an improvement to another technology or technical field. Accordingly, the claims are not integrated into a practical application.
Applicant further argues that the claims also recite an improvement to the claimed interface and web page by providing interactive user elements. For example, amended claim 1 recites, in part, "transmitting the at least one top alternative product for display to the user on an interactive web page of a user device, the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product." At least this additional claim element provides discrete, specific ways of how the interface elements functionalities are related to the overall system and demonstrate an improvement to computer functionality (Remarks, page 14).
Examiner respectfully disagrees. Merely displayed data and interactive user elements doesn’t improve the actual interface technology, it implements or applies the abstract idea on a generic computing hardware (i.e. a generic interface) and does no more than generally link the use of the judicial exception to a particular technological environment or field of use. Accordingly, the amended claims are not integrated into a practical application.
Applicant further argues that, under step 2B, Applicant submits that the claims are not "directed to" an abstract idea and qualify as patent-eligible subject matter, and further analysis under Step 2B is not required. Even if, arguendo, the claims are found to be directed to an abstract idea, Applicant submits that, under Step 2B, the claims, as a whole, recite "significantly more" than the alleged abstract idea. Applicant's amended claim 1 recites significantly more than the alleged abstract idea. For example, existing systems are limited due to being unable to generate a recommendation in response to a user query. See paragraph [003] of the Specification. For example, the claims recite, ""based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query." At least this claim element addresses the ability to generate recommendations even if there is not an exact match to a user query. This provides an improvement to the functioning of computer systems in the same vein as cases such as DOR Holdings, in that these elements enable functionality that was not possible with prior systems (Remarks, page 15).
Examiner respectfully disagrees. Improving the ability to generate recommendations even if there is not an exact match to a user query is not a technical improvement (see MPEP 2106.05(a)), rather, it is a business improvement. Additionally, in DDR, the claims overcome a problem or propose a solution “specifically arising in the realm of computer [technology]” DDR Holdings, 773 F.3d at 1257. Unlike DDR, there is no recited technical improvement. Accordingly, the amended claims do not amount to significantly more than an abstract idea.
Applicant further argues that the claims also recite a technical improvement by identifying specific improvements to the way that the top alternative product is displayed to the user on a web page. For example, amended claim 1 recites, "transmitting the at least one top alternative product for display to the user on an interactive web page of a user device, the web page including interactive user elements regarding the top alternative product, including at least one selectable element, a picture of the top alternative product, and an ordered list of sellers associated with the top alternative product" (Remarks, pages 15-16).
Examiner respectfully disagrees. Merely displayed data and interactive user elements doesn’t improve the actual interface technology, it implements or applies the abstract idea on a generic computing hardware (i.e. a generic interface) and does no more than generally link the use of the judicial exception to a particular technological environment or field of use. Accordingly, the amended claims do not amount to significantly more than an abstract idea.
Applicant further argues that independent claim 1 is directed to patent eligible subject matter. Independent claims 11 and 20, although different in scope, include similar elements and are also directed to patent-eligible subject matter for the same reasons. Claims 1-10, and 12-19 each depend from independent claims 1 or 11 and are patent-eligible at least due to their dependency (Remarks, page 16).
Examiner respectfully disagrees. As detailed in response to the arguments above, claim 1 is ineligible. Accordingly, claims 11 and 20, as well as, dependent claims 2-10 and 12-19 are ineligible.
Rejections under 35 U.S.C. §103
Applicant argues that Garg does not teach "based on the determination of no matches, using a machine learning algorithm to determine at least one top alternative product based on the plurality of search attributes associated with the product search query" as Garg discloses the use of a "threshold" applied to the "relevance score," to filter out results. However, the "threshold," is applied to the "base set of results," in Garg rather than used as the initial filter as the Office suggests. In contrast, the present application first determines there are "no matches," and then uses an algorithm to determine "at least one top alternative product." Therefore, Garg does not disclose at least this claim element (Remarks, pages 16-17).
Garg teaches querying an ecommerce web site for attributes matching particular attributes [i.e. based on the plurality of search attributes associated with the product search query] in a search query and determining the top similar items [i.e. based on the determination of no matches, determine at least one top alternative product] based on similarity scores determined by a continually learning feedback-loop [i.e. using a machine learning algorithm to] (see Garg, Col. 11 Ln. 45-52, Col. 11 Ln. 53-65, and Col. 6 Ln. 37-42). Examiner notes that if the relevance score is below 200, then there isn’t an exact match and this is in accordance with paragraph [004] of Applicant spec. Paragraph [004] of Applicant spec indicates that “if no exact match exists based on the product of interest, the methods or systems would automatically identify the likely product categories associated with the user's product of interest.” Additionally, the threshold is applied to the base results before they are sent to the continually learning feedback-loop ranking systems. Accordingly, the cited references teach the amended claims.
Applicant further argues that Garg does not disclose the amended claim element of, "performing an iterative match between the product search query and a product index data set using iterative string matching and a set of key features." Emphasis added. Garg does not disclose the use of "iterative string matching," nor the use of both "iterative string matching and a set of key features," to perform the iterative match (Remarks, page 17).
Applicant’s argument has been considered but is moot because this arguments does not apply to the current combination of references being used to teach this amended feature.
Applicant further argues that Although of different scope, independent claims 11 and 20 include elements similar to those discussed above for claim 1. For at least the same reasons presented above in connection with the patentability of claim 1, independent claims 11 and 20 are patentable and, thus, allowable. Dependent claims 2-10 and 12-19 are also allowable at least by virtue of their dependence from one of independent claims 1 or 11 . Therefore, withdrawal of the rejections of claims 1-20 under 35 U.S.C.§ 103 is respectfully requested (Remarks, pages 17-18).
Examiner respectfully disagrees. As detailed above, the cited references teach claim 1. Accordingly, independent claims 11 and 20, as well as, dependent claims 2-9 and 12-18 are not allowable. Examiner notes that claims 10 and 19 are allowable over the prior art as detailed above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
-Berkowitz et al. (US 2012/0265744 A1) teaches an intelligent system that searches for an item.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ARIELLE E WEINER whose telephone number is (571)272-9007. The examiner can normally be reached M-F 8:30-5:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Maria-Teresa (Marissa) Thein can be reached at 571-272-6764. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ARIELLE E WEINER/ Primary Examiner, Art Unit 3689