DETAILED ACTION
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 4/1/26 has been entered.
Status of the Claims
This office action is submitted in response to the RCE filed on 4/1/26.
Examiner notes that this application is a continuation of 16/456830, which was abandoned on 5/14/24.
Examiner further notes Applicant’s priority date of 6/28/19, which stems from the aforementioned parent application.
Claims 4, 7, 14, and 17 have been canceled.
Claims 1, 11, and 20 have been amended.
Therefore, claims 1-3, 5-6, 8-13, 15-16, and 18-20 are currently pending and have been examined.
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 Objections
Claims 1, 11, and 20 are objected to because of the following informalities:
Claims 1, 11, and 20, as amended, recite that the predicted number of user interactions is based on interactions between pairs of selected items in the plurality of selected items (claim 1) or between each combination of two of the selected items in the plurality of selected items (claims 11 and 20) and each of the plurality of recommended items related to at least one of the plurality of selected items. The amended limitation retains the sentence structure of the previous claim language, in which the recited interactions ran between (i) each of the plurality of selected items and (ii) each of the plurality of recommended items. As amended, the grammatical relationship between the recited combinations of selected items and the recited recommended items is unclear. Specifically, the limitation is amenable to a first reading in which the recited interactions occur between each two-item combination of selected items, taken as a unit, and each of the recommended items, and a second reading in which the recited interactions occur among the pairwise combinations of the selected items themselves, with the recommended items being related to at least one of the selected items. Appropriate correction is required. See MPEP § 608.01(m); 37 C.F.R. § 1.75.
Claims 1, 11, and 20 are further objected to because each recites a graphical user interface of a user device in the receiving limitation but subsequently recites the graphic user interface in the transmitting limitation. Consistent terminology is required. Appropriate correction is required.
Claim Interpretation
For purposes of examination, and consistent with the original disclosure, the limitation the predicted number of user interactions is based on interactions between pairs of selected items in the plurality of selected items [claims 11 and 20: between each combination of two of the selected items in the plurality of selected items] and each of the plurality of recommended items related to at least one of the plurality of selected items is being interpreted in accordance with its broadest reasonable interpretation consistent with the specification, pursuant to MPEP § 2111, as requiring that (a) the predicted number of user interactions is based on pairwise relationships among the items in the plurality of selected items, consistent with the transition matrix described with reference to FIG. 3A of the instant application, wherein a probability is generated for each combination of two of the items, and (b) the plurality of recommended items are related to at least one of the plurality of selected items, consistent with the per-anchor-item recommended item lists described in the instant specification. The Examiner notes that the specification does not describe an interaction occurring between a two-item combination of selected items, taken as a unit, and a recommended item. Accordingly, the alternative reading identified in the objection above is not consistent with the specification, and the claims have not been examined under that reading.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-3, 5-6, 8-13, 15-16, and 18-20 are rejected under 35 U.S.C. 112(a) as failing to comply with the written description requirement. The claims contain subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, at the time the application was filed, had possession of the claimed invention.
Independent claims 1, 11, and 20 recite moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance of the plurality of recommended content items. The originally filed disclosure does not support this limitation.
The specification expressly characterizes FIG. 5A as illustrating recommended item data identifying item advertisements that may be selected by the item advertisement computing device 102. Elements 520, 530, and 540 of FIG. 5A are described as first, second, and third item recommendation items data, respectively. FIG. 5A therefore depicts stored data structures, namely, per-anchor-item lists of digital advertisements, and does not depict a graphical user interface. FIG. 5B depicts the online shopping cart webpage 502 on which a selection of digital advertisements is displayed. The specification describes selecting digital advertisements from the recommendation items data for display on the webpage, and describes the order in which the selected advertisements are displayed. The specification does not describe any advertisement being displayed at a first position on a graphical user interface and subsequently moved to a second position on the graphical user interface, whether based on an order of relevance or otherwise.
Further, the specification states that the list may rank recommended items in an order of relevance, for example, that recommended items that are purchased more often with the anchor item may appear before recommended items that are purchased less often with the anchor item. This passage describes the ordering of a stored list of recommended items for each anchor item, which may be stored in database 116. It does not describe, and does not reasonably convey, the movement or repositioning of displayed content items to positions on a graphical user interface as a function of relevance.
Finally, the specification describes, with reference to FIG. 5C, that upon selection of the additional recommendation icon 514, the current digital advertisements may be removed from the online shopping cart webpage 502 and the next digital advertisements of the digital advertisement list 550 may be displayed. This passage describes the wholesale replacement of one set of displayed advertisements with a subsequent set in response to a user selection. It does not describe moving displayed content items to positions on the graphical user interface based on an order of relevance.
Therefore, because the originally filed disclosure describes only the selection of recommended items, the ordering of stored lists of recommended items, the display of selected advertisements, and the replacement of displayed advertisements in response to user input, and does not describe moving displayed content items to positions on a graphical user interface based on an order of relevance, the limitation moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance of the plurality of recommended content items lacks written description support.
Claims 2-3, 5-6, 8-10, 12-13, 15-16, and 18-19 depend from claims 1 and 11, incorporate the above limitation, and are rejected on the same basis.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 5-6, 10-13, 15-16, and 18-19 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Claim 5 recites the limitation the subset of recommended items. There is insufficient antecedent basis for this limitation in the claim, as neither claim 5 nor claim 1, from which it depends, previously recites a subset of recommended items.
Claim 6 recites the limitations the stochastic process and the transition matrix probabilities. There is insufficient antecedent basis for these limitations in the claim, as neither claim 6 nor claim 1, from which it depends, previously recites a stochastic process or transition matrix probabilities.
Claim 10 recites the limitation the stochastic process. There is insufficient antecedent basis for this limitation in the claim, as neither claim 10 nor claim 1, from which it depends, previously recites a stochastic process.
Claim 11 recites the limitation determining, by the processor. There is insufficient antecedent basis for this limitation in the claim, as claim 11 does not previously recite a processor. Claims 12-13, 15-16, and 18-19 depend from claim 11, incorporate the above limitation, and are rejected on the same basis.
Claim 15 recites the limitation the subset of recommended items. There is insufficient antecedent basis for this limitation in the claim, as neither claim 15 nor claim 11, from which it depends, previously recites a subset of recommended items.
Claim 16 recites the limitations the stochastic process and the transition matrix probabilities. There is insufficient antecedent basis for these limitations in the claim, as neither claim 16 nor claim 11, from which it depends, previously recites a stochastic process or transition matrix probabilities.
Claim 19 recites the limitation the stochastic process. There is insufficient antecedent basis for this limitation in the claim, as neither claim 19 nor claim 11, from which it depends, previously recites a stochastic process.
Considerations Under 35 USC § 101
Independent claims 1, 11, and 20 recite, in part, the collection of user activity data comprising a plurality of selected items, the determination of a predicted number of user interactions that a plurality of recommended content items receive based on pairwise relationships among the selected items, the ranking of the selected items based on a measure, and the selection of recommended content items related to at least one of the selected items. These limitations describe the collection and analysis of data to generate targeted item recommendations, which is a judicial exception pursuant to MPEP § 2106.04(a)(2), as it falls within the mental processes and certain methods of organizing human activity (commercial interactions/advertising) groupings of abstract ideas. The amendments to the independent claims further defining the pairwise relationships upon which the predicted number of user interactions is based serve to narrow the abstract idea itself and do not alter this characterization.
However, the independent claims additionally recite automatically transmit[ting] for display the plurality of recommended content items on the graphical user interface of the user device and moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance of the plurality of recommended content items. This limitation remains in the claims as amended. As set forth in the previous office action, this GUI relocation behavior constitutes an additional element analogous to that of Example 37 of the Office’s Subject Matter Eligibility Examples, in which a claim reciting automatically moving the most used icons to a position on the GUI closest to the start icon based on a determined amount of use was deemed eligible. Considered as a whole, the claims integrate the judicial exception into a practical application under Step 2A, Prong Two, because the additional elements recite a specific manner of automatically displaying and positioning recommendations on a computer interface based on their relevance, which provides a specific improvement over prior systems, resulting in an improved user interface for electronic devices. See MPEP § 2106.04(d).
Therefore, a rejection under 35 U.S.C. § 101 is not appropriate at this time. However, as set forth in the rejection under 35 U.S.C. § 112(a) above, the limitation upon which this conclusion rests lacks written description support in the originally filed disclosure. Applicant is again advised that amendment of the claims to remove or materially alter this limitation in order to comply with 35 U.S.C. § 112(a) will likely necessitate a rejection under 35 U.S.C. § 101 in a future office action.
Claim Rejections - 35 USC § 103
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.
Claims 1-3, 5-6, 8-13, 15-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Jacobi et al. (US 6,317,722 B1) in view of Gramatica (US 10,025,862 B2), and in further view of Aravamudan et al. (US 8,375,069 B2).
Claims 1, 11, and 20: Jacobi discloses a system, method, and non-transitory computer readable medium comprising:
a processor; and a non-transitory memory storing instructions that, when executed, cause the processor to (FIG. 1; col. 7, ll. 30-46; col. 8, ll. 9-13. The web site 30 includes a web server application 32 that processes HTTP requests received over the Internet from user computers, and the recommendation and shopping cart processes are described as code modules that are executed by a computer system to perform particular tasks, which corresponds to a processor executing instructions stored in memory.):
receive, via a graphical user interface of a user device, a user request to display a plurality of recommended content items (FIG. 6; col. 9, ll. 1-11; col. 10, ll. 1-9. A browser window is shown in which the user selects a hyperlink such as “Instant Book Recommendations” or initiates display of the shopping cart page, and in response a list of recommended items is generated and displayed on the same web page as the shopping cart contents.);
determine, by the processor, in real-time the predicted number of user interactions the plurality of recommended content items receive based on activity data of the user, the activity data comprise a plurality of selected items (FIG. 1; FIG. 2; FIG. 7; col. 4, ll. 24-38; col. 5, ll. 19-28; col. 6, ll. 39-67; col. 7, ll. 1-24; col. 9, ll. 29-55. Items currently in the user’s shopping cart, which is a data structure that keeps track of items that have been selected by the user, are identified as items of known interest, the current and/or recent shopping cart contents are used as inputs to the recommendation service, and in the shopping-cart implementation the recommendations are generated in real-time when the user initiates a display of the shopping cart and are displayed on the same web page as the shopping cart contents.);
each of the plurality of recommended items related to at least one of the plurality of selected items (Abstract; col. 2, ll. 52-67; col. 3, ll. 17-34; FIG. 1; FIG. 3. An additional item is selected for inclusion in the list based in part upon whether that item is related to more than one of the items in the user’s shopping cart, using a similar-items table mapping items to lists of similar items based on purchase-history correlations.); and
automatically transmit for display the plurality of recommended content items on the graphical user interface of the user device (FIG. 2; FIG. 6; FIG. 7; col. 6, ll. 39-67; col. 9, ll. 1-11. The recommendations are generated and displayed automatically when the user views the shopping cart contents, with the recommendations list incorporated into the web page returned to the user’s browser for display together with the shopping cart contents.).
While Jacobi bases recommendations on purchase histories, item-to-item commonality indices, and shopping cart contents, and sorts the resulting recommendation list from highest to lowest score (FIG. 2, steps 82, 88; FIG. 7, steps 286-288), Jacobi does not explicitly disclose wherein the specific criteria is a predicted number of user interactions the plurality of recommended content items receive, nor that the predicted number of user interactions is based on interactions between pairs of selected items in the plurality of selected items [claims 11 and 20: between each combination of two of the selected items in the plurality of selected items] and each of the plurality of recommended items related to at least one of the plurality of selected items, and the plurality of the selected items are ranked based on a measure.
Gramatica, however, discloses a method wherein the specific criteria is a predicted number of user interactions the plurality of recommended content items receive and the predicted number of user interactions is based on interactions between pairs of selected items in the plurality of selected items [claims 11 and 20: between each combination of two of the selected items in the plurality of selected items] and each of the plurality of recommended items related to at least one of the plurality of selected items, and the plurality of the selected items are ranked based on a measure (Abstract; FIG. 5; col. 9, ll. 35-42; col. 9, ll. 56-62; col. 10, ll. 4-11; col. 10, ll. 29-62; col. 10, l. 67 through col. 11, l. 5. A weighted information network is constructed in which the aij element of the similarity matrix is the weight of the link connecting node i and node j, a transfer matrix P is defined whose elements are conditional transition probabilities p(i|j) between nodes derived from co-occurrence dependencies, and a random walker navigates the graph, jumping from a vertex to a neighbor with probability proportional to the dependency between the two nodes. The random walk distance between any two nodes is computed as dRW = Σ[1/(I−B(j))], where B(j) is an N×N matrix identical to the transfer matrix P, such that the distance for any pair of nodes is computed over the pairwise transition probabilities among all combinations of two nodes in the network; the probability of any given path is likewise the product of the pairwise transition probabilities p(i←v1)p(v1←v2)…p(vk←j) between the nodes traversed, extracted as the strongest dependency chain between any two or more nodes. The random walk distances are calculated for each pair of nodes, yielding a complete graph in which each node is connected to every node and each pair is connected by a link whose weight is the random walk distance between the two nodes, and the RW distances constitute an implicit ranking measure for each couple of distinct nodes. These pairwise distance scores accordingly rank nodes according to the strength of their connection to one or more query nodes, which corresponds to predicted likelihoods of interaction between the selected items and the candidate recommended items and to a ranking based on a measure.).
Therefore, it would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Jacobi’s shopping-cart-based recommendation service to employ Gramatica’s random-walk analysis over a weighted item network, with the user’s cart items as query nodes, as the specific criterion for ranking and selecting the recommended items. One would have been motivated to do this because Jacobi already computes pairwise item-to-item commonality indices from purchase histories, and Gramatica’s transfer-matrix random walk provides a principled probabilistic mechanism for converting exactly such pairwise item dependencies into predicted interaction scores for ranking candidate items.
Further, while Jacobi describes sorting recommendation lists by score to determine which items to recommend and in what order to present them in a list, Jacobi does not explicitly disclose moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance of the plurality of recommended content items.
Aravamudan, however, discloses a method for moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance of the plurality of recommended content items (Abstract; col. 2, ll. 15-35; col. 7, ll. 35-67; col. 8, ll. 1-31; FIG. 10. Per-user relevance weights are associated with content items, those weights are adjusted based on user navigation and selection actions, and a subset of content items and content groups is selected and presented to the user ordered by the adjusted associated relevance weights; the navigation hierarchy is reordered to match the user’s action behavior, items with higher relevance weights are presented higher in the results, and items can be brought closer to the user’s locus of relevance or hoisted to the root, which corresponds to moving items to different positions in the interface based on relevance ordering.).
Therefore, it would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to combine these features of Aravamudan with those of Jacobi and Gramatica. One would have been motivated to do this in order to improve the presentation of Jacobi’s recommended items by arranging and dynamically repositioning them on the GUI so that more relevant recommendations appear in more prominent positions, thereby reducing user effort to find desired content.
Claims 2 and 12: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the instructions further cause the processor to: determine that a customer has interacted with an item from the plurality of selected items based on receiving one or more interactions from the user interface; and receive, from a web server, an online cart activity message identifying one or more cart items that have been added to an online digital cart (FIG. 1; col. 5, ll. 19-28; col. 7, ll. 45-64; col. 8, ll. 8-19. Users add and remove items to/from a personal shopping cart, a shopping cart process within external components 40 adds and removes items from the users’ personal shopping carts based on the actions of the respective users, and a user profiles database 38 stores the current contents of the user’s personal shopping cart(s) and recent shopping cart contents, which necessarily occurs in response to user interactions on the site and corresponds to receiving cart-activity data indicating items added to the cart.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claims 3 and 13: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the instructions further cause the processor to: receive, from a web server, a subsequent online cart activity message identifying one or more cart items that have been removed from the online digital cart (FIG. 1; col. 5, ll. 19-28; col. 7, ll. 54-58; col. 8, ll. 8-19. The shopping cart feature allows users to add and remove items to/from a personal shopping cart, the shopping cart process adds and removes items from the users’ personal shopping carts, and the user profiles database maintains recent shopping cart contents, defined as items that were recently removed from the shopping cart(s) without being purchased, which corresponds to subsequent online cart activity messages indicating items removed from the cart.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claims 5 and 15: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the instructions further cause the processor to: generate an item advertisement recommendation message identifying recommended items to be included in one of the subset of recommended items (FIG. 1; FIG. 2; FIG. 6; FIG. 7; col. 4, ll. 24-38; col. 6, ll. 39-67; col. 8, ll. 20-25; col. 8, ll. 55-63; col. 9, ll. 1-11. The recommendation process 52 generates a recommendations list, and recommendations generated by the recommendation services are returned to the web server 32, which incorporates the recommendations into personalized web pages transmitted to users, which requires producing a data structure or message that identifies recommended items for inclusion in a displayed recommendation subset.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claims 6 and 16: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Gramatica further discloses a system wherein executing the stochastic process comprises executing a random walk algorithm based on the transition matrix probabilities (Abstract; FIG. 5; col. 9, ll. 1-5; col. 10, ll. 29-53. The knowledge graph is crawled by a stochastic process, e.g., a random-walk analyser, wherein a transfer matrix P of transition probabilities p(i|j) is defined from the link weights and the random walk is executed over this matrix, with the walker moving between vertices with probability given by the transition probabilities, to compute the random walk distances and rankings.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claim 8: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the plurality of selected items includes a first item and a second item, and the instructions further cause the processor to generate a first set of digital advertisements for the first item and a second set of digital advertisements for the second item, the first item being different than the second item (FIG. 1; FIG. 2; FIG. 3; FIG. 7; col. 3, ll. 17-34; col. 4, ll. 24-38; col. 8, ll. 1-24. A similar-items list is retrieved from the table for each shopping cart item that is a popular item (FIG. 7, step 282) and these per-item lists are used to generate recommendations, corresponding to separate sets of recommended items for each selected cart item which, when displayed, function as per-item recommendation sets for different items.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claims 9 and 18: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the instructions further cause the processor to: determine a maximum number of recommended items for each item of the plurality of selected items (FIG. 2; FIG. 3; FIG. 7; col. 4, ll. 24-38; col. 8, ll. 1-24. Each similar-items list 64 is truncated (FIG. 3, step 116) and the recommendation process recommends the top M items from the list (FIG. 2, step 94; FIG. 7, step 294), which establishes and applies maximum counts for recommended items per base item.).
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Claims 10 and 19: The Jacobi/Gramatica/Aravamudan combination discloses those limitations cited above.
Jacobi further discloses a system wherein the instructions further cause the processor to: generate a first set of recommended items for a first item of the plurality of selected items, the first set including a predetermined number of recommended items and cause the user device to display the first set of recommended items (FIG. 2; FIG. 5; FIG. 6; FIG. 7; col. 6, ll. 39-67; col. 9, ll. 1-11. A list of recommended items is generated per base item, the top M items from the list are recommended, and the recommended items are displayed on a web page, where the list contains a predetermined number M of recommended items shown to the user.).
Jacobi does not explicitly disclose generate a second set of recommended items for the first item of the plurality of selected items, the second set including the predetermined number of recommended items, displaying the first set in response to execution of the stochastic process, or cause the user device to display the second set of recommended items in response to completion of displaying the first set of recommended items and in response to an indication of the customer’s input on the user interface.
Aravamudan, however, discloses a system for generat[ing] a second set of recommended items for the first item of the plurality of selected items, the second set including the predetermined number of recommended items and caus[ing] the user device to display the second set of recommended items in response to completion of displaying the first set of recommended items and in response to an indication of the customer’s input on the user interface (Abstract; col. 2, ll. 15-35; col. 4, ll. 1-30; col. 7, ll. 35-67; FIG. 10. A subset of items ordered by relevance weights is presented, those weights are updated based on user navigation and selection actions, and new subsets of items are then selected and presented in response to continued user input, corresponding to displaying subsequent sets of items after an initial display when the user continues to interact with the interface.). As to displaying the first set in response to execution of the stochastic process, the execution of the stochastic process is taught by Gramatica as set forth above with respect to claims 6 and 16, and in the combination the display of Jacobi’s recommendations follows the execution of Gramatica’s random walk ranking.
The rationale for combining Jacobi, Gramatica, and Aravamudan is articulated above and reincorporated herein by reference.
Other Relevant Prior Art
The following references are not cited above, but are nevertheless deemed to be relevant to Applicant’s disclosures:
Agarwal et al. (20240029139), directed to a method and apparatus for item selection.
Joshi et al. (8417559), directed to assortment planning based on demand transfer between products.
Chen et al. (20110184806), directed to probabilistic recommendation of an item.
Das et al. (20210263939), directed to transition regularized matrix factorization for sequential recommendation.
Gunawardana et al. (20090006290), directed to training random walks over absorbing graphs.
Response to Arguments
Rejection under 35 U.S.C. § 112(a)
Applicant argues (Remarks, pp. 8-9) that the rejection of claims 1, 11, and 20 under 35 U.S.C. § 112(a) should be withdrawn because (i) the claims do not recite "dynamically moving" or "repositioning," but rather recite moving the plurality of recommended content items to positions on the graphical user interface based on an order of relevance; (ii) drawings alone may satisfy the written description requirement, citing Vas-Cath Inc. v. Mahurkar, 935 F.2d 1555 (Fed. Cir. 1991) and MPEP § 2163(I), and FIGS. 5A and 5B show recommended item AD1,3 (526), arranged adjacent AD1,2 (524) in FIG. 5A, being moved to a different location on the GUI in FIG. 5B; and (iii) paragraph 35 of the application as filed states that the list may rank recommended items in an order of relevance. These arguments are not persuasive.
As an initial matter, the Examiner does not dispute the legal framework Applicant advances. Drawings alone may provide written description support, possession may be shown in a variety of ways, and there is no in haec verba requirement. See Vas-Cath, 935 F.2d at 1564; MPEP § 2163(I). The dispositive question, however, is not whether drawings can convey possession as a matter of law, but whether these drawings, read in light of the specification's own description of them, reasonably convey to a person of ordinary skill in the art that the inventors possessed moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance. They do not.
Applicant's showing rests on a comparison of the position of element 526 in FIG. 5A with its position in FIG. 5B. But the specification expressly characterizes FIG. 5A as illustrating recommended item data, and describes elements 520, 530, and 540 as first, second, and third item recommendation items data, respectively. FIG. 5A depicts stored data structures, namely per-anchor-item lists of digital advertisements. It does not depict a graphical user interface, and no advertisement is displayed to any user in FIG. 5A. FIG. 5B, by contrast, depicts the online shopping cart webpage 502 on which a selection of digital advertisements is displayed. A difference between the position of an element within a stored data structure and the position of that element upon display does not depict, and would not reasonably convey to a person of ordinary skill in the art, the movement of a displayed content item from one position to another position on a graphical user interface. At most, the figures convey that advertisements are selected from stored recommendation data and displayed in a selection order that differs from the order in which they are stored, which is precisely the selection-and-display subject matter the Examiner acknowledges is supported. Because element 526 is never depicted as displayed on the GUI at a first position, the figures cannot show it being moved to a second position, whether based on an order of relevance or otherwise.
Applicant's reliance on paragraph 35 fares no better. The cited passage states that "[t]he list may rank recommended items in an order of relevance" and that the list of recommended items for each anchor item may be stored in database 116. This passage describes the ordering of a stored list, not the position of displayed content items on a graphical user interface, and certainly not the movement of displayed content items to positions on the interface as a function of relevance. Indeed, the passage confirms the Examiner's characterization of the disclosure: the specification is directed to which items are selected for recommendation and the order in which they are listed, not to GUI-level rearrangement.
Finally, Applicant's observation that the claims do not recite the words "dynamically moving" or "repositioning" does not advance Applicant's position. The claims recite moving the plurality of recommended content items to positions on the graphic user interface based on an order of relevance. Under the broadest reasonable interpretation, moving a content item to a position on a graphical user interface presupposes both a starting position and an ending position on the interface; an item that is rendered once at a single position has been displayed, not moved. The Examiner's use of the word "dynamically" in the previous office action was descriptive of this behavior, which is the same behavior exhibited in Example 37 of the Subject Matter Eligibility Examples, wherein icons having existing positions on the GUI are automatically relocated to positions closest to the start icon based on a determined amount of use. Notably, Applicant's own remarks embrace this parallel, asserting that the claims as amended provide a specific improvement "like Example 37" (Remarks, p. 9); Applicant cannot rely upon the Example 37 repositioning behavior to support eligibility while simultaneously disclaiming that same behavior to avoid the written description requirement. The originally filed disclosure describes selecting items for display, displaying the selected items in an order, and, with reference to FIG. 5C, replacing one set of displayed advertisements with the next set in response to selection of the additional recommendation icon 514. In each case, an item is rendered at a position and remains there until it is removed; no embodiment describes a displayed item occupying a first position on the interface and subsequently being relocated to a second position, based on an order of relevance or otherwise. The words the Examiner used to describe the missing subject matter do not alter the subject matter that is missing.
The Examiner further notes that the moving limitation cannot reasonably be construed as requiring merely that the recommended content items be displayed in an order of relevance. As set forth in the Considerations Under 35 U.S.C. § 101 above, the eligibility of the claims rests on the moving limitation constituting a specific manner of automatically repositioning displayed recommendations on the interface, analogous to Example 37 of the Subject Matter Eligibility Examples, that integrates the recited judicial exception into a practical application. A construction under which the limitation requires no more than displaying the results of the recited ranking in a ranked order would not provide such an integration, as it would amount to mere display of the results of the abstract idea, and would necessitate a rejection under 35 U.S.C. § 101. The claim as written is therefore supported only under a construction the specification does not describe, and describable only under a construction the claim does not recite.
Accordingly, Applicant's arguments have been fully considered but are not persuasive. The rejection of claims 1-3, 5-6, 8-13, 15-16, and 18-20 under 35 U.S.C. § 112(a) is therefore maintained.
Rejections under 35 U.S.C. § 103
Applicant argues (Remarks, pp. 10-11) that claim 1 as amended is patentable over the cited references because Jacobi does not describe the predicted number of user interactions is based on interactions between pairs of selected items in the plurality of selected items and each of the plurality of recommended items related to at least one of the plurality of selected items, asserting that Jacobi describes providing recommendations based on the content of the entire cart, different weighing factors based on ratings, and recency of purchase. Applicant further asserts that "Gramatica and Aravamudan are not relied upon to address this feature," and that claims 11 and 20, and the dependent claims, are patentable for at least similar reasons. These arguments are not persuasive.
Applicant's argument attacks Jacobi individually for a limitation that the rejection does not map to Jacobi. In the previous office action, as in the rejection set forth above, the limitation concerning the predicted number of user interactions, including the basis for that prediction and the ranking of the selected items based on a measure, is expressly mapped to Gramatica, not to Jacobi. The rejection relies upon Jacobi for the shopping-cart-based recommendation environment, including the identification of cart items as items of known interest and the generation and display of recommendations related to those items, and relies upon Gramatica for the probabilistic prediction and ranking criterion applied within that environment. Applicant's assertion that Gramatica is not relied upon to address this feature is therefore incorrect as a characterization of the rejection of record. Nonobviousness cannot be established by attacking references individually where the rejection is based upon a combination of references. In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986); MPEP § 2145(IV). Because Applicant's arguments do not address Gramatica's teachings, they do not identify any deficiency in the rejection.
Moreover, to the extent the amended claims are argued to require that the prediction be based on pairwise relationships among the selected items, Gramatica teaches this feature, as set forth in detail in the rejection above. Gramatica defines a similarity matrix whose aij element is the weight of the link connecting node i and node j, defines a transfer matrix P whose elements are conditional transition probabilities between nodes, and computes a random walk distance for each pair of nodes, yielding a complete graph in which each node is connected to every node and each pair is connected by a link whose weight is the random walk distance between the two nodes. The random walk distance for any pair of nodes is computed over the transfer matrix as a whole, such that the resulting scores are based on the pairwise transition probabilities among all combinations of two nodes in the network, including each combination of two of the query nodes, and Gramatica expressly teaches extracting the strongest dependency chain between any two or more nodes. In the combination set forth in the rejection, the items in the user's shopping cart of Jacobi serve as the query nodes, and the prediction and ranking are accordingly based on the pairwise dependencies among the selected items and the candidate recommended items, consistent with the claim interpretation set forth above.
With respect to claims 11 and 20, Applicant presents no separate argument beyond the assertion that these claims are patentable for at least similar reasons as claim 1, and with respect to the dependent claims, Applicant relies solely on their dependency. These arguments are unpersuasive for the reasons set forth above with respect to claim 1.
Accordingly, Applicant's arguments have been fully considered but are not persuasive. The rejection of claims 1-3, 5-6, 8-13, 15-16, and 18-20 under 35 U.S.C. § 103 is therefore maintained.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHRISTOPHER BUSCH whose telephone number is (571)270-7953. The examiner can normally be reached M-F 10-7.
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, Waseem Ashraf can be reached at 571-270-3948. 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.
/CHRISTOPHER C BUSCH/Examiner, Art Unit 3621
Prosecution Insights