DETAILED ACTION
This communication is in response to the Application filed on 2/28/2024. Claims 1-20 are pending and have been examined.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 13, 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 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1, 13, and 17 recite A method comprising: determining, based on a search result selected by a user, using a [first large language model (LLM)], a first set of categories, a first set of keywords, a second set of categories, and a second set of keywords, wherein: the first set of categories is a predetermined set of categories, and the first set of keywords corresponds to at least one category of the first set of categories; the second set of categories is based on user information associated with the user, and the second set of keywords corresponds to at least one category of the second set of categories; retrieving a set of digital content items, wherein a first digital content item of the set of digital content items is retrieved based on a first category of the first set of categories, and wherein a second digital content item of the set of digital content items is retrieved based on a first category of the second set of categories; generating, using a [second LLM], a summary of one or more digital content items of the set of digital content items; and causing the summary to be [displayed on a device].
The limitations in these claims, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. This process can be explained as a mental process by using the example of someone asking a librarian for help with book recommendations. The user selecting a search result is a person asking the librarian for recommendations based on a book they selected. The librarian could use their knowledge of the book to determine categories and keywords associated with it. The librarian would have some predetermined categories such as popular titles/genres that they recommend to everyone. The librarian also uses information about the person such as their age, gender, and any interests they may already know they have. The librarian could then select all the books they would recommend as use their prior knowledge to summarize the books they're recommending. They could then present this to the person by writing it down on a sheet of paper. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of a first and second large language model and a device for displaying. The first LLM is described as being any pretrained LLM in paragraph 49 of the specification. In paragraph 62 of the specification, it is said that the second LLM can be the same as the first LLM. The device for displaying is described in paragraph 164 of the specification with a generic description of the component. Furthermore, the device is considered pre-solution and post-solution activity that is merely used as a data gathering step and data presentment step for the user. Claim 13 specifically lists additional components a processor and a memory. The processor for displaying is described in paragraph 159 of the specification with a generic description of the component. The memory for displaying is described in paragraph 158 of the specification with a generic description of the component. Claim 17 specifically lists additional component of a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium for displaying is described in paragraph 170 of the specification with a generic description of the component. Accordingly, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claims are not patent eligible.
Claims 2, 14, and 18 recite wherein a third digital content item of the set of digital content items is retrieved based on a keyword of the first set of keywords and wherein a fourth digital content item of the set of digital content items is retrieved based on a keyword of the second set of keywords, wherein the summary is a first summary of the first digital content item and the third digital content item, further comprising: generating, using the second LLM, a second summary of the second digital content item and the fourth digital content item.
The limitations in these claims, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could use their knowledge of the books to find predetermined and user specific keywords that could help them make recommendations. They could then create a first summary of book recommendations based only on the predetermined recommendations they make to every person. Then they could create a second summary of book recommendations based on what that specific person might like. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claims do not recite any additional elements that were not present in the independent claim. Accordingly, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claims are not patent eligible.
Claims 3, 15, and 19 recite wherein the search result is displayed on the device using a webpage, and wherein causing the summary to be displayed further comprises augmenting the webpage with the summary.
The limitations in these claims, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian can display the recommendations to the user using a sheet of paper, this summary could be displayed/augment a webpage by the sheet of paper being placed next to a webpage. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claims do not recite any additional elements that were not present in the independent claim. Accordingly, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claims are not patent eligible.
Claims 4, 16, and 20 recite wherein causing the summary to be displayed on the device is in response to the selection of the search result by the user.
The limitations in these claims, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian can display the recommendations to the user using a sheet of paper. As explained in the independent claim the device is performing post-solution activity of merely presenting the information produced by the method. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claims do not recite any additional elements that were not present in the independent claim. Accordingly, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claims are not patent eligible.
Claim 5 recites ranking the set of digital content items based on: a relevance of a digital content item of the set of digital content items to the at least one category of the first set of categories, a relevance of the digital content item of the set of digital content items to at least one keyword of the first set of keywords, a relevance of the digital content item of the set of digital content items to the at least one category of the second set of categories, or a relevance of the digital content item of the set of digital content items to at least one keyword from the second set of keywords.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could rank their book recommendations based on how relevant they are to the predetermined or user specific categories or keywords. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 6 recites wherein generating, using the second LLM, the summary of the set of digital content items further comprises: selecting a number of ranked digital content items; and generating, using the second LLM, a summary of the number of ranked digital content items.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could order their recommendation summarization based on a ranked relevance such as putting their strongest recommendations first. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 7 recites wherein the first LLM receives a first prompt and the second LLM receives a second prompt, wherein the first prompt is different from the second prompt.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, in this case the prompts are the librarian handling two different tasks mentally. The first one being noting all the things they know about the book the person presented to them. The second being making a summarization of books they'd recommend to the person. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 8 recites wherein the summary comprises a reference identifier associated with a digital content item of the set of digital content items.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could write a reference for each book such as where to find it in the library. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 9 recites wherein the first digital content item of the set of digital content items is retrieved from a first database and the second digital content item of the set of digital content items is retrieved from a second database.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could find recommendations from multiple databases such as one being their own memory of the books and another being a reference journal of all the books they have in the library. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 10 recites determining, based on the search result selected by the user, using the first LLM, a third set of categories and a third set of keywords, wherein the third set of categories is based on content of the search result selected by the user, and the third set of keywords corresponds to at least one category of the third set of categories.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could find categories and keywords that are specific to the book the person wants recommendations based on. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 11 recites obtaining content of the search result selected by the user using metadata of the search result selected by the user; and obtaining the user information associated with the user using metadata associated with the user.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, metadata could be information about the book such as the author. There could also be personal metadata such as if this person has read books from the author before. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim 12 recites wherein determining, using the first LLM, the first set of categories, the first set of keywords, the second set of categories, and the second set of keywords, is further based on the content of the search result selected by the user and the user information associated with the user.
The limitations in this claim, as drafted, are a process that, under broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. From the independent claim example, the librarian could find categories and keywords that are specific to the book the person wants recommendations based on. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. The claim does not recite any additional elements that were not present in the independent claim. Accordingly, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim Rejections - 35 USC § 102
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.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-8 and 10-20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US Patent Application Publication US 20240289407 A1 (Rofouei et al.).
Regarding Claims 1, 13, and 17, Rofouei et al. teaches A method comprising:
Alternatively, claim 13 states, A system comprising: at least one processor; and at least one memory device coupled to the at least one processor, wherein the at least one memory device comprises instructions that, when executed by the at least one processor, cause the at least one processor to perform at least one operation comprising:
Alternatively, claim 17 states, A non-transitory computer-readable storage medium comprising instructions that, when executed by at least one processor, cause the at least one processor to perform at least one operation comprising:
(In some implementations, a method may be implemented by one or more processors during a search session of a user and may include:) (Paragraph 19).
(These software modules are generally executed by processor 1014 alone or in combination with other processors. Memory 1025 used in the storage subsystem 1024 can include a number of memories including a main random access memory (RAM) 1030 for storage of instructions and data during program execution and a read only memory (ROM) 1032 in which fixed instructions are stored. A file storage subsystem 1026 can provide persistent storage for program and data files, and may include a hard disk drive, a floppy disk drive along with associated removable media, a CD-ROM drive, an optical drive, or removable media cartridges) (Paragraph 221).
determining, based on a search result selected by a user, using a first large language model (LLM),
(In some implementations disclosed herein, multiple LLMs are utilized in series in generating an NL based summary responsive to a query. As one example, first LLM(s) can be used to select passages from a set of search result document(s) (e.g., utilized in block 260A of method 200 of FIG. 2). Second LLM(s) can then be used to generate a summary for each of the passage(s) selected utilizing the first LLM(s) (e.g., utilized in block 260A1 of method 200 of FIG. 2). Further, third LLM(s) can then generate an overall NL based summary based on the individual passage summaries generated using the second LLM(s) (e.g., utilized in block 260B of method 200 of FIG. 2).) (Paragraph 150).
(The “can benefit from ambient generative summarization” classification may cause LLM selection engine 132 to select and trigger LLM response generation engine 136 to process digital content (e.g., a SRD) from the SRP that is consumed by user (result/action) using ambient generative LLM(s) 872C. The ambient generative summarization of a particular digital content or domain in which the user is engaged (e.g., access, clicks through, consumes) may be generated, e.g., by LLM response generation engine 136 using ambient generative LLM(s) 872C, based primarily on current query 870 and one or more results and/or actions performed by the user. These results and/or actions may include the user reading and/or interactive with a particular SRD (e.g., a webpage).) (Paragraph 177).
The system Rofouei et al presents summarizes search results for a user using multiple LLMs. Search result documents (SRDs) and prior search result pages (SRPs) are used throughout the presented methods to preform summaries. One particular capability of the invention is create a summary based on user selecting an SRD that was part of a SRP, effectively choosing a result from a group of search results.
a first set of categories, a first set of keywords, a second set of categories, and a second set of keywords, wherein:
(In some implementations, block 254 includes sub-block 254A in which selecting the set of query-responsive SRDs can be based on query-dependent measure(s), query-independent measure(s), and/or user-dependent measure(s) for the query-responsive SRDs.) (Paragraph 66).
(Query-independent measures for a query-responsive SRD can include, for example, a selection rate of the query-responsive search result document for multiple queries, a trustworthiness measure for the query-responsive search result document (e.g., one generated based on an author thereof, a domain thereof, and/or inbound link(s) thereto), an overall popularity measure for the query-responsive search result document, and/or a freshness measure that reflects recency of creation or updating of the query-responsive search result document.) (Paragraph 68).
(User-dependent measures for a query-responsive SRD can be based on, for example, relation of the query-responsive search result document to: attributes of a user profile for the query; recent queries at the client device or via the user profile; and/or recent non-query interactions at the client device or via the user profile. For example, if a user profile indicates a user is a movie buff, then user-dependent measure(s) for an SRD pertaining to a movie can result in the SRD more likely being selected for inclusion in the set than if the user profile did not indicate the user is a movie buff.) (Paragraph 69).
(It is noted that, by considering query-dependent and/or user-specific measure(s) in selecting the set at block 254A, different sets can be determined for different submissions of the same query. For example, using a locality measure can lead to a first set of query-specific SRDs for the query of “history of Louisville” submitted from Louisville, Kentucky and a distinct second set of query-specific SRDs for the query of “history of Louisville” submitted from Louisville, Colorado. As described herein, differing sets of SRD(s) will result in differing generated NL based summaries.) (Paragraph 70).
The SRD selection engine in Fig. 1 uses various measures determined from the input. These query-independent measures and the user-dependent measures act as a first and second set of categories determined from an input where the values within these measures act as keywords.
the first set of categories is a predetermined set of categories, and the first set of keywords corresponds to at least one category of the first set of categories;
(Query-independent measures for a query-responsive SRD can include, for example, a selection rate of the query-responsive search result document for multiple queries, a trustworthiness measure for the query-responsive search result document (e.g., one generated based on an author thereof, a domain thereof, and/or inbound link(s) thereto), an overall popularity measure for the query-responsive search result document, and/or a freshness measure that reflects recency of creation or updating of the query-responsive search result document.) (Paragraph 68).
This set of measures used to retrieve documents is considered predetermined as they are independent of the input and the user and thus would be the same for every input. Categories such as trustworthiness and popularity are considered and keywords associated with them would be the values associated with the trustworthiness or popularity.
the second set of categories is based on user information associated with the user, and the second set of keywords corresponds to at least one category of the second set of categories;
(User-dependent measures for a query-responsive SRD can be based on, for example, relation of the query-responsive search result document to: attributes of a user profile for the query; recent queries at the client device or via the user profile; and/or recent non-query interactions at the client device or via the user profile. For example, if a user profile indicates a user is a movie buff, then user-dependent measure(s) for an SRD pertaining to a movie can result in the SRD more likely being selected for inclusion in the set than if the user profile did not indicate the user is a movie buff.) (Paragraph 69).
This set of measures used to retrieve documents is considered associated with the user as they are user-dependent. Categories such as attributes of the profile and recent queries are considered and keywords associated with them would be the actual attributes and recent searches within those categories.
retrieving a set of digital content items, wherein a first digital content item of the set of digital content items is retrieved based on a first category of the first set of categories, and wherein a second digital content item of the set of digital content items is retrieved based on a first category of the second set of categories;
(In some implementations, block 254 includes sub-block 254A in which selecting the set of query-responsive SRDs can be based on query-dependent measure(s), query-independent measure(s), and/or user-dependent measure(s) for the query-responsive SRDs. In some implementations, the system includes a search system that optionally generates one or more of such measures.) (Paragraph 66).
(In some implementations, data from SRP(s) returned by search system(s) may be provided as additional input to ambient generative LLM(s) 872C. For instance, suppose the user is on a particular company's web page and asks, “What competing model is similar to this?” The content of the webpage and content from the SRP (e.g., passages, content from other web page(s) linked to in the SRP) that includes information about the competing model may be included as input for ambient generative LLM(s) 872C.) (Paragraph 79).
Both the category/keyword measures and the the users selection of an SRD from the SRP are considered in the process of selecting documents that will be used as input to the summary LLM
generating, using a second LLM, a summary of one or more digital content items of the set of digital content items;
(At block 260, the system generates an NL based summary based on processing, using an LLM, corresponding content from each of the SRD(s) of the set determined in block(s) 256, 258, 259, and/or 260. For example, if five search result documents are selected for the set, a corresponding portion of content from each of the five can be processed using the LLM to generate the NL based summary.) (Paragraph 77).
(In some implementations disclosed herein, multiple LLMs are utilized in series in generating an NL based summary responsive to a query. As one example, first LLM(s) can be used to select passages from a set of search result document(s) (e.g., utilized in block 260A of method 200 of FIG. 2). Second LLM(s) can then be used to generate a summary for each of the passage(s) selected utilizing the first LLM(s) (e.g., utilized in block 260A1 of method 200 of FIG. 2). Further, third LLM(s) can then generate an overall NL based summary based on the individual passage summaries generated using the second LLM(s) (e.g., utilized in block 260B of method 200 of FIG. 2).) (Paragraph 150).
The selected documents are used to generate a summary via a second LLM.
and causing the summary to be displayed on a device.
(At block 262, the system causes the NL based summary, generated at block 260, to be rendered in response to the query. For example, the system can cause the NL based summary to be rendered graphically in an interface of an application of a client device via which the query was submitted.) (Paragraph 84).
The generated summary is displayed to the user.
Regarding Claims 2, 14, and 18, Rofouei et al. teaches the method of claims 1, 13, and 17,
wherein a third digital content item of the set of digital content items is retrieved based on a keyword of the first set of keywords and wherein a fourth digital content item of the set of digital content items is retrieved based on a keyword of the second set of keywords,
(Query-independent measures for a query-responsive SRD can include, for example, a selection rate of the query-responsive search result document for multiple queries, a trustworthiness measure for the query-responsive search result document (e.g., one generated based on an author thereof, a domain thereof, and/or inbound link(s) thereto), an overall popularity measure for the query-responsive search result document, and/or a freshness measure that reflects recency of creation or updating of the query-responsive search result document.) (Paragraph 68).
(User-dependent measures for a query-responsive SRD can be based on, for example, relation of the query-responsive search result document to: attributes of a user profile for the query; recent queries at the client device or via the user profile; and/or recent non-query interactions at the client device or via the user profile. For example, if a user profile indicates a user is a movie buff, then user-dependent measure(s) for an SRD pertaining to a movie can result in the SRD more likely being selected for inclusion in the set than if the user profile did not indicate the user is a movie buff.) (Paragraph 69).
As explained above the categories are represented by the measures above where the “keywords” would be the results to these measures.
wherein the summary is a first summary of the first digital content item and the third digital content item, further comprising: generating, using the second LLM, a second summary of the second digital content item and the fourth digital content item.
(It is noted that, by considering query-dependent and/or user-specific measure(s) in selecting the set at block 254A, different sets can be determined for different submissions of the same query. For example, using a locality measure can lead to a first set of query-specific SRDs for the query of “history of Louisville” submitted from Louisville, Kentucky and a distinct second set of query-specific SRDs for the query of “history of Louisville” submitted from Louisville, Colorado. As described herein, differing sets of SRD(s) will result in differing generated NL based summaries. For example, differing sets of SRDs will result in differing content, from the SRDs of the respective set, being processed using the LLM in generating the respective NL based summary. Accordingly, differing generated NL based summaries will be provided for different submissions of the same query.) (Paragraph 70).
(For example, each of the multiple LLMS can be utilized to generate a corresponding candidate NL based summary, but only one of the candidate NL based summaries selected for use (e.g., for rendering in response to the query). For instance, one can be selected for use based on it (a) being similar to the greatest quantity of other candidate NL based summaries, (b) being similar to at least a threshold quantity of other candidate NL based summaries, (c) lacking certain content (e.g., certain term(s)), (d) including certain content (e.g., certain term(s)), (d) having the highest language model score, (e) having a language model score that satisfies a threshold, and/or (f) having or lacking other feature(s).) (Paragraph 147)
The measures can result in different SRDs being found and thus different summaries being generated. The above example shows locations (category) for Louisville, Kentucky and Colorado (keywords) which are measures used to determine SRDs specific to those measures (in this case the SRDs represent a second and fourth content item). Multiple summaries can be generated both with and without these terms/features/measures and they are eventually condensed into one output summary.
Regarding Claims 3, 15, and 19, Rofouei et al. teaches the method of claims 1, 13, and 17,
wherein the search result is displayed on the device using a webpage, and wherein causing the summary to be displayed further comprises augmenting the webpage with the summary.
(At block 458, the system monitors for interaction(s) with any of the search result document(s) that are responsive to the query. For example, the system can determine an interaction with a search result document based on determining a selection of the corresponding link) (Paragraph 117).
(FIG. 7A depicts an example client device 710 with a display 780 rendering, in response to a query 782, a graphical interface that includes an example NL based summary 784 and additional example search results 788 that are responsive to a query. In the NL based summary 784, there are three linkified portions, each indicated by underlining and a source identifier (S1, S2, S3) provided immediately following the linkified portions.) (Paragraph 144).
The user interface shown in Fig. 7 is web-based utilizing links. The users input can modify this interface and result in an updated summary being displayed as shown in Fig.4
Regarding Claims 4, 16, and 20, Rofouei et al. teaches the method of claims 1, 13, and 17,
wherein causing the summary to be displayed on the device is in response to the selection of the search result by the user.
(The system proceeds to block 460 if an interaction with search result document(s) is determined at block 458. At block 460, the system generates a revised NL based summary based on processing revised input using the LLM or an additional LLM. The revised input reflects the occurrence of the interaction(s) with the search result document(s), and is revised relative to the input that is processed, using the LLM, in block 454.) (Paragraph 118).
Fig. 4 shows the process of a revised summary being rendered based on a user selecting a search result.
Regarding Claim 5, Rofouei et al. teaches the method of claim 1,
ranking the set of digital content items based on:
(At block 254, the system selects one or more query-responsive search result documents (SRDs), that are responsive to the query of block 252, for inclusion in a set. For example, the system can select, for inclusion in the set, a subset of query-responsive SRDs that the system and/or a separate search system have identified as responsive to the query. For instance, the system can select the top N (e.g., 2, 3, or other quantity) query-responsive SRDs as determined by a search system or can select up to N query-responsive SRDs that have feature(s), as determined by the system, that satisfy one or more criteria.) (Paragraph 65)
(In some implementations, block 254 includes sub-block 254A in which selecting the set of query-responsive SRDs can be based on query-dependent measure(s), query-independent measure(s), and/or user-dependent measure(s) for the query-responsive SRDs.) (Paragraph 66).
The SRDs are ranked by the search system as well as indirectly ranked by how well they fit the measures used for collecting them.
a relevance of a digital content item of the set of digital content items to the at least one category of the first set of categories,
a relevance of the digital content item of the set of digital content items to at least one keyword of the first set of keywords,
(Query-independent measures for a query-responsive SRD can include, for example, a selection rate of the query-responsive search result document for multiple queries, a trustworthiness measure for the query-responsive search result document (e.g., one generated based on an author thereof, a domain thereof, and/or inbound link(s) thereto), an overall popularity measure for the query-responsive search result document, and/or a freshness measure that reflects recency of creation or updating of the query-responsive search result document.) (Paragraph 68).
The SRDs can be “ranked” according to a score such as selection rate where the score would correspond to the relevance it has to the predetermined categories.
a relevance of the digital content item of the set of digital content items to the at least one category of the second set of categories,
or a relevance of the digital content item of the set of digital content items to at least one keyword from the second set of keywords.
(User-dependent measures for a query-responsive SRD can be based on, for example, relation of the query-responsive search result document to: attributes of a user profile for the query; recent queries at the client device or via the user profile; and/or recent non-query interactions at the client device or via the user profile. For example, if a user profile indicates a user is a movie buff, then user-dependent measure(s) for an SRD pertaining to a movie can result in the SRD more likely being selected for inclusion in the set than if the user profile did not indicate the user is a movie buff.) (Paragraph 69).
The SRD can be determined based on its relation to the user dependent metrics mentioned previously to encompass both category and keyword.
Regarding Claim 6, Rofouei et al. teaches the method of claim 5,
wherein generating, using the second LLM, the summary of the set of digital content items further comprises: selecting a number of ranked digital content items;
(At block 254, the system selects one or more query-responsive search result documents (SRDs), that are responsive to the query of block 252, for inclusion in a set. For example, the system can select, for inclusion in the set, a subset of query-responsive SRDs that the system and/or a separate search system have identified as responsive to the query. For instance, the system can select the top N (e.g., 2, 3, or other quantity) query-responsive SRDs as determined by a search system or can select up to N query-responsive SRDs that have feature(s), as determined by the system, that satisfy one or more criteria.) (Paragraph 65)
The SRDs found can be ranked by the search system to restrict the number given to the LLM for summarization.
and generating, using the second LLM, a summary of the number of ranked digital content items.
(At block 262, the system causes the NL based summary, generated at block 260, to be rendered in response to the query. For example, the system can cause the NL based summary to be rendered graphically in an interface of an application of a client device via which the query was submitted.) (Paragraph 84).
The LLM generates a summary from the selection of SRD’s
Regarding Claim 7, Rofouei et al. teaches the method of claim 1,
wherein the first LLM receives a first prompt and the second LLM receives a second prompt,
wherein the first prompt is different from the second prompt.
(In some implementations disclosed herein, multiple LLMs are utilized in series in generating an NL based summary responsive to a query. As one example, first LLM(s) can be used to select passages from a set of search result document(s) (e.g., utilized in block 260A of method 200 of FIG. 2). Second LLM(s) can then be used to generate a summary for each of the passage(s) selected utilizing the first LLM(s) (e.g., utilized in block 260A1 of method 200 of FIG. 2). Further, third LLM(s) can then generate an overall NL based summary based on the individual passage summaries generated using the second LLM(s) (e.g., utilized in block 260B of method 200 of FIG. 2).) (Paragraph 150).
Different LLMs are used for the SRD gathering and summary generation thus necessitating unique prompts being used for each.
Regarding Claim 8, Rofouei et al. teaches the method of claim 1,
wherein the summary comprises a reference identifier associated with a digital content item of the set of digital content items.
(In some implementations, at further sub-block 260A1 the system can additionally or alternatively include, as part of the content, a source identifier of the SRD. For example, the source identifier can be a token included at the beginning and/or the end of the content. The token can be unique relative to other source identifier(s) for other SRD(s) of the set.) (Paragraph 82).
The summary includes source identifiers for the SRDs used in the summary.
Regarding Claim 10, Rofouei et al. teaches the method of claim 1,
determining, based on the search result selected by the user, using the first LLM, a third set of categories and a third set of keywords,
(Query-dependent measures for a query-responsive SRD can include, for example, a positional ranking of the query-responsive search result document and for the query, a selection rate of the query-responsive search result document and for the query, a locality measure that is based on an origination location of the query and a location corresponding to the query-responsive search result document, and/or a language measure that is based on a language of the query and a language corresponding to the query-responsive search result document.) (Paragraph 67).
(The revised input can reflect familiarity with content of the search result document(s) interacted with and, as a result, the revised NL based summary will be updated in view of that familiarity. For example, the revised NL based summary can omit content of the search result document(s) interacted with (whereas the NL based summary of block 454 included it), or the revised NL based summary can include a more in depth discussion of the content of the search result document(s) interacted with (whereas the NL based summary of block 454 included only a high level overview of the content) the occurrence of the interaction(s) with the search result document(s).) (Paragraph 118).
(The “can benefit from ambient generative summarization” classification may cause LLM selection engine 132 to select and trigger LLM response generation engine 136 to process digital content (e.g., a SRD) from the SRP that is consumed by user (result/action) using ambient generative LLM(s) 872C. The ambient generative summarization of a particular digital content or domain in which the user is engaged (e.g., access, clicks through, consumes) may be generated, e.g., by LLM response generation engine 136 using ambient generative LLM(s) 872C, based primarily on current query 870 and one or more results and/or actions performed by the user. These results and/or actions may include the user reading and/or interactive with a particular SRD (e.g., a webpage).) (Paragraph 177).
New measures from and SRP related to the SRD the user interacted with are considered after the user’s interaction. Categories and keywords in this case would be equivalent to the query-dependent measures as they measures specifically tied to the users input.
wherein the third set of categories is based on content of the search result selected by the user, and the third set of keywords corresponds to at least one category of the third set of categories.
(Query-dependent measures for a query-responsive SRD can include, for example, a positional ranking of the query-responsive search result document and for the query, a selection rate of the query-responsive search result document and for the query, a locality measure that is based on an origination location of the query and a location corresponding to the query-responsive search result document, and/or a language measure that is based on a language of the query and a language corresponding to the query-responsive search result document.) (Paragraph 67).
(At sub-block 4606, the system generates the revised NL based summary using an additional LLM, relative to the one used in block 454, that is fine-tuned based on a prompt that reflects familiarity with content of the SRD(s) interacted with. For example, the fine-tuned LLM model can be trained to receive known content that the user already knows, followed by (e.g., after a delimiter) additional content to be summarized in view of the known content. For instance, the known content portion of the input can be based on content from the given search result.) (Paragraph 121)
(In some implementations, the generation of ambient generative content by LLM response generation engine 136 may include grounding, recitation, and/or attribute checking from the SRD itself and/or from data associated with the SRD, such as its metadata. In some implementations, LLM selection engine 132 may trigger ambient generative summarization based on explicit input from the user.) (Paragraph 178).
The third set of categories could be represented by the query-dependent measures the system uses. This includes categories such as a positional ranking in the results and a selection rate of the result where the keywords would be the values assigned to these categories. Furthermore, it is mentioned how the summary can be altered based on the selection of a document so it may not include information from that SRD in the updated summary and data associated with the SRD selected by the user becomes part of response generation engine.
Regarding Claim 11, Rofouei et al. teaches the method of claim 1,
obtaining content of the search result selected by the user using metadata of the search result selected by the user;
(Another downstream LLM may include an SRP summarization LLM. A SRP summarization LLM may be trained to summarize SRPs, including passages of text that are rendered in proximity to the links, context extracted from the linked-to documents, titles of the documents, other metadata associated with the documents, and so forth.) (Paragraph 13).
The system also utilizes prior search result document (SRP) information including the metadata from those documents.
and obtaining the user information associated with the user using metadata associated with the user.
(As shown by the arrow from current output 874 to user model 871, in various implementations, user state 871 may be updated based on current output 874, as well as any accompanying metadata. For example, summarizations (SRD or SRP) and/or creatives generated by LLM response generation engine 136 may be added to user state 871 in their native/raw form, or may be encoded into semantically-rich embeddings, bags of words (e.g., using techniques such as TF-IDF), etc., and incorporated with user state 871. Consequently, these data may influence various information generated by various components of FIG. 8 during the next turn or iteration of the generative companion.) (Paragraph 194).
This metadata can be added to a user profile (See Fig. 8) and can influence future iterations of the method.
Regarding Claim 12, Rofouei et al. teaches the method of claim 11,
wherein determining, using the first LLM, the first set of categories, the first set of keywords, the second set of categories, and the second set of keywords, is further based on the content of the search result selected by the user and the user information associated with the user.
(Another downstream LLM may include an SRP summarization LLM. A SRP summarization LLM may be trained to summarize SRPs, including passages of text that are rendered in proximity to the links, context extracted from the linked-to documents, titles of the documents, other metadata associated with the documents, and so forth.) (Paragraph 13).
(User-dependent measures for a query-responsive SRD can be based on, for example, relation of the query-responsive search result document to: attributes of a user profile for the query; recent queries at the client device or via the user profile; and/or recent non-query interactions at the client device or via the user profile. For example, if a user profile indicates a user is a movie buff, then user-dependent measure(s) for an SRD pertaining to a movie can result in the SRD more likely being selected for inclusion in the set than if the user profile did not indicate the user is a movie buff.) (Paragraph 69).
(The “can benefit from ambient generative summarization” classification may cause LLM selection engine 132 to select and trigger LLM response generation engine 136 to process digital content (e.g., a SRD) from the SRP that is consumed by user (result/action) using ambient generative LLM(s) 872C. The ambient generative summarization of a particular digital content or domain in which the user is engaged (e.g., access, clicks through, consumes) may be generated, e.g., by LLM response generation engine 136 using ambient generative LLM(s) 872C, based primarily on current query 870 and one or more results and/or actions performed by the user. These results and/or actions may include the user reading and/or interactive with a particular SRD (e.g., a webpage).) (Paragraph 177).
The system identifies user information as part of the initial measures for selecting documents. Furthermore, the SRPs provide context to the users prior searches and upon selection of an SRD within the SRP a new summary is generated containing the added context and metadata from the SRP.
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.
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.
Claim 9 are rejected under 35 U.S.C. 103 as being unpatentable over US Patent Application Publication US 20240289407 A1 (Rofouei et al.). in view of US Patent Publication US 11755651 B2 (Hopkins et al.).
Regarding Claim 9, Rofouei et al. teaches the system of claim 1.
Rofouei et al. does not explicitly teach: wherein the first digital content item of the set of digital content items is retrieved from a first database and the second digital content item of the set of digital content items is retrieved from a second database.
However, Hopkins et al. teaches wherein the first digital content item of the set of digital content items is retrieved from a first database and the second digital content item of the set of digital content items is retrieved from a second database.
(The process for determining a decomposition category can include (1) mapping at least one query fragment in the one or more query fragments to at least one candidate decomposition category based at least in part on one or more category indicators corresponding to one or more decomposition categories and one or more category-comparison) (Col. 4, Lines 26-35).
(Each of the sets of category indicators includes one or more words and/or phrases (or one or more combinations of words and/or phrases), collectively called “characteristic phrases.” Each individual or group of these characteristic phrases is associated with one or more specific categories from Box 210 (which is the same as Box 104).) (Col. 9, Lines 29-34).
(Collectively or individually, the original user-entered search term from Box 101, the core search term, if any, from Boxes 503A and 607A, and/or the identified search filters, if any, from Boxes 503B and 607B can be used to execute a search in a plurality of ways and in a plurality of modes against a plurality of data or information repositories. As illustrated by Boxes 106A and 106B in FIG. 1, two of these are: (1) searching one or more databases of products, services, news or information items, websites, web applications, mobile applications, images, videos, and/or other pieces of data for instances included therein that match one or more specific search terms, mediated and constrained by the extracted search filters and filter values, such that the search is carried out against only those database items whose specified search-filter values match some or all of the corresponding search-filter values of the user-entered search term;) (Col. 14, Lines 59-67 - Col. 15, Lines 1-8).
In Hopkins et al. a search query entered by the user is broken down into query fragments which are then made into query filters. The query filters are then used to search through multiple databases that are relevant to that filter. The initial queries are broken up into categories and characteristic phrases that are associated with those categories thus making it the same method as that used to retrieve the digital content items.
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify the search result summarization method as taught by Rofouei et al. to search separate databases for the different categories extracted from the query as taught by Hopkins et al. This would have been an obvious improvement to ensure that all relevant databases are included in the search and less relevant data is excluded (Hopkins et al. Col. 1, Lines 64-67 to Col. 2, Lines 1-18).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NICHOLAS DANIEL LOWEN whose telephone number is (571)272-5828. The examiner can normally be reached Mon-Fri 8:00am - 4:00pm.
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, Paras D Shah can be reached at (571) 270-1650. 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.
/NICHOLAS D LOWEN/Examiner, Art Unit 2653
/Paras D Shah/Supervisory Patent Examiner, Art Unit 2653
11/07/2025
Prosecution Insights