DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
2. This Office Action is issued in response to the Appeal Brief Conference decision on July 20, 2026.
3. Claims 1-20 are pending, of which claims 1, 19, and 20 are in independent form.
Information Disclosure Statement
4. The information disclosure statement (IDS) submitted on April 30, 2026. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the IDS is being considered by the examiner.
Response to Arguments
5. Applicant’s argument, see “§ 112 Rejections” filed on July 20, 2026 have been carefully considered. Based on claim amendments and persuasive remarks filed, the claim rejections have been withdrawn.6. Applicant’s argument, see “§ 101 rejection”, filed on July 20, 2026 have been carefully considered and are not persuasive.. I. Applicant argues that the claims address technical problems in existing document-search systems
7. Applicant’s Argument: Existing similarity-scoring systems perform syntactic comparisons between documents and therefore may identify documents using similar terminology without identifying documents that are semantically related to the problem expressed in the source document. Applicant further argues that existing systems fail to intelligently map information from particular fields of a source document to corresponding fields of repository documents. Applicant relies on Specification ¶ [0020] for these alleged deficiencies. Applicant additionally relies on Specification ¶ [0024] and argues that conventional systems generally permit comparison only between documents sharing the same schema, thereby limiting the number of repository documents that can accurately be considered responsive. According to Applicant, the disclosed system overcomes this limitation by permitting semantic comparisons between documents having different schemas.
Examiner’s Response: Examiner has carefully considered the argument, but respectfully disagrees. The fact that the Specification identifies shortcomings associated with certain conventional document-comparison techniques does not, by itself, establish that independent claim 1 recites a technological improvement to computer functionality. The eligibility inquiry considers the limitations actually required by the claim and whether the claim, as a whole, integrates the identified judicial exception into a practical application. In particular, Applicant's asserted technological improvement relies substantially on the ability to perform comparisons across different or heterogeneous document schemas. However, that limitation is not required by independent claim 1. Independent claim 1 recites a first schema associated with the first document and a second schema associated with the documents of the document repository, but does not require those schemas to differ. Rather, dependent claim 2 separately recites: “wherein the first schema is different from the second schema.” Accordingly, the asserted improvement directed specifically to overcoming the inability of conventional systems to compare documents having different schemas is not comparable with the scope of independent claim 1. At the breadth of claim 1, the claim encompasses obtaining information associated with documents, accessing a configuration defining how searches are to be generated, generating searches according to that configuration, obtaining results, and using the results to determine similarity. Although these operations may improve the relevance or usefulness of the information produced by the search, claim 1 does not require the particular cross-schema implementation upon which Applicant relies in describing the alleged technological improvement. Therefore, the technical problems identified in Specification ¶¶ [0020] and [0024] do not establish that the full scope of independent claim 1 itself recites the asserted improvement to computer or database-search technology.
II. Applicant argues that the claimed configuration and targeted queries provide a technical solution and technological improvement
8. Applicant’s Argument: Applicant relies on Specification ¶ [0021] and contends that the claimed architecture solves the identified technical problems through defined configurations that instruct the system how to generate intelligent queries linking the meaning of information in a source document to information contained in repository documents. Applicant further argues that the disclosed technique avoids inefficient comparisons between all fields and instead generates targeted queries between particular fields, thereby improving the accuracy and efficiency of similarity processing.
Examiner’s Response: Examiner has carefully considered the argument, but respectfully disagrees. Applicant's characterization of the disclosed embodiments is broader than the technological implementation actually required by independent claim 1. Although claim 1 recites accessing a configuration defining how queries are generated and generating queries based upon the configuration, the claim does not require a particular technological mechanism for implementing those operations. For example, claim 1 does not require a particular database indexing structure, search algorithm, data structure, memory arrangement, processor architecture, or other modification to the underlying operation of the computer or database system. Nor does independent claim 1 require the particular field-selection, weighting, indexing, or other implementation details relied upon by Applicant in characterizing the disclosed embodiments as avoiding inefficient or brute-force comparisons. Instead, at the level of generality recited in independent claim 1, the configuration provides rules or instructions governing how information is searched and compared. The processor applies those rules to formulate queries, obtains information responsive to the queries, and evaluates the resulting information to determine similarity. Thus, even accepting that the disclosed embodiments may produce more relevant similarity results or reduce unnecessary comparisons in particular implementations, the independent claim does not require the specific implementation details responsible for those asserted benefits. An improvement described in the Specification does not establish eligibility where the claim is broad enough to encompass implementations that do not require that asserted technological improvement. Accordingly, Applicant has not persuasively established that the claimed configuration and query-generation limitations, as actually recited in independent claim 1, constitute an improvement to the functioning of the computer, database architecture, indexing mechanism, or other underlying technology.
III. Applicant argues that the claims themselves recite the technological solution that achieves the asserted improvement
9. Applicant’s Argument: Applicant contends that independent claim 1, and corresponding independent claims 19 and 20, do not merely recite the desired result of identifying similar documents. Rather, Applicant argues that the claims expressly recite the schema-specific configuration and targeted query-generation operations that produce the technological improvement described in the Specification.
Examiner’s Response: Examiner has carefully considered the argument, but respectfully disagrees. The Examiner acknowledges that eligibility must be evaluated based upon the claim as a whole rather than merely upon the stated objective of identifying similar documents. However, considering the claim as a whole does not demonstrate that the asserted improvement is required by the claimed subject matter. As discussed above, independent claim 1 does not require the first and second schemas to be different. The alleged technological advantage associated with searching across heterogeneous schemas therefore is not required by the independent claim. Similarly, although Applicant characterizes the claimed queries as “targeted” queries that eliminate inefficient comparisons between all fields, independent claim 1 broadly requires accessing a configuration defining how to generate queries and generating the queries based upon that configuration. The claim does not require a particular algorithm or technological implementation that produces the asserted efficiency improvement.
Rather, considered as an ordered combination, the claimed operations concern obtaining document-related information, using predetermined instructions to formulate searches of information, obtaining results of those searches, and evaluating or combining those results to determine similarity. The recited processor and memory perform those information-processing operations according to the claimed rules, but the claim does not recite a particular improvement to the operation of the processor, memory, database architecture, indexing mechanism, or underlying computer functionality itself. Applicant's reliance upon the Specification does not alter this conclusion. The Specification may be relevant in determining whether Applicant has identified an asserted technological improvement and in understanding the claimed invention. However, the asserted improvement must be reflected in the limitations required by the claim. Here, the Specification describes embodiments involving comparisons between documents having different schemas and targeted comparisons between particular fields, while independent claim 1 does not require all of the limitations upon which Applicant relies in asserting those technological advantages. Accordingly, Applicant has not persuasively established that independent claim 1 requires the asserted improvement involving heterogeneous schemas or the particular efficiency-enhancing implementation described in the Specification.
IV. Applicant's Step 2A, Prong Two argument does not establish integration into a practical application
10. Applicant’s Argument: Applicant further argues that, under Step 2A, Prong Two, the foregoing features integrate the alleged abstract idea into a practical application because the claims solve a specific technical problem and improve computer functionality.
Examiner’s Response: Examiner has carefully considered the argument, but respectfully disagrees. The mere use of a processor, memory, document repository, configuration, and generated queries does not establish integration into a practical application where those components are used to carry out the recited information-processing operations without a claimed improvement to their underlying technological operation. Here, the additional elements implement the claimed process of obtaining information, generating searches according to predetermined rules, retrieving information, and determining similarity. Applicant has identified advantages that may arise from particular embodiments described in the Specification. However, independent claim 1 does not require the heterogeneous-schema relationship expressly relied upon by Applicant, nor does it require the particular field-selection, weighting, indexing, or other technological implementation asserted to provide improved efficiency. Therefore, when independent claim 1 is considered as a whole, Applicant's arguments do not establish that the judicial exception is integrated into a practical application under Step 2A, Prong Two. Accordingly, the rejection of independent claim 1 under 35 U.S.C. § 101 is maintained.
11. Applicant’s arguments, see “§ 103 Rejection”, filed on July 20, 2026 have been carefully considered and are not persuasive..
I. Neither Baker nor Zhang discloses accessing a configuration for a first schema defining how to generate queries into a collection having a second schema
12. Applicant argues: Baker merely generates synthetic query suggestions by extracting text from HTML structures of documents on the same website and therefore does not disclose accessing a configuration for a first schema defining how to generate, from a first document, a plurality of queries into a collection of documents having a second schema. Applicant further argues that Zhang does not cure this deficiency because Zhang operates on term-frequency lists rather than schema-based configurations.
Examiner’s Response: Applicant's arguments have been considered but are not persuasive. As an initial matter, Applicant's argument relies upon a limitation that is not recited in independent claim 1. Applicant repeatedly characterizes the claimed first and second schemas as “different” schemas. However, claim 1 merely recites a first document having a first schema and a collection of documents having a second schema. It is dependent claim 2 that expressly requires “the first schema [to be] different from the second schema.” Accordingly, the fact that Baker may apply its templates to documents having similar structures does not distinguish the full scope of claim 1. Furthermore, Baker expressly teaches structured documents having a defined structure. Baker explains: “The structure can include markups of the document.” [col. 7, line 1-2]. Baker further teaches that the structure may include metadata, HTML tag structure, text fragments associated with the tag structure, and hierarchical structures in which tags are nested. [col. 7, lines 7-8] Thus, Baker expressly recognizes and processes the structural organization of the documents rather than merely treating the documents as undifferentiated text. Baker further teaches generating a query template from that document structure: “Each query template 216 includes one or more of the identified structures 214 that are then used to produce candidate synthetic queries” [col. 7, lines 27-30] and subsequently applies those templates to structured documents to extract words, phrases, or text segments matching the structures defined by the templates. Even more directly, Baker states: “The query template includes a generative rule to be used in generating candidate synthetic queries.” [col. 16, lines 4-5]. When the generative rule is applied to other structured documents, it causes generation of one or more candidate synthetic queries based upon the embedded coding fragment included in the template. Baker thereafter generates the candidate synthetic queries by “applying the generative rules in the query templates to the other structured documents.” [col. 16, lines 25-27]. Thus, Baker's query template/generative rule provides the claimed configuration because it defines instructions governing how information is extracted from structured documents to generate queries. Baker additionally demonstrates that application of a query template can generate a plurality of queries. In its example, application of a template to another structured document produces four candidate synthetic queries: “joseph heller catch,” “joseph heller short,” “joseph heller stories,” and “joseph heller short stories.” [col. 11, lines 49-51].Accordingly, Applicant's characterization of Baker as merely extracting text from webpages does not account for Baker's express disclosure of structured documents, query templates containing structural constraints and generative rules, and generation of multiple queries according to those rules. Therefore, Applicant's arguments do not overcome the rejection.
II. Neither Baker nor Zhang discloses combining results of executed plurality of queries into similarity scores for the first document.
13. Applicant’s Argument: Zhang merely calculates cosine similarity between term-frequency vectors and therefore does not execute a plurality of queries generated from a schema configuration or combine the execution results of those queries into document similarity scores.
Examiner’s Response: Applicant's arguments have been considered but are not persuasive because they address the references individually and additionally impose limitations that are not expressly recited in claim 1. First, claim 1 does not expressly require “executing” the plurality of queries. Claim 1 requires “generating the plurality of queries based on the configuration” and “combining results of the plurality of queries into similarity scores for the first document.” The additional requirement of “executing the plurality of queries on the collection of documents” is expressly introduced in dependent claim 16. Moreover, Baker does disclose obtaining results associated with the generated queries. After generating candidate synthetic queries, Baker teaches: “The system measures (610) a performance in a search operation of each of the candidate synthetic queries.” [col. 16, lines 28-29]. Baker further states that the system may “calculate a query performance score for the candidate synthetic query.” [col. 16, lines 30-32] Baker elsewhere explains that the score may be an information retrieval (IR) score calculated in relation to the structured document. [col. 2, lines 66-67]. Zhang expressly supplies the document-similarity calculation relied upon in the combination. Zhang obtains multiple indexing terms from a document and, “For each index terms, a frequency of the term,” after which it identifies document categories similar to the document based upon the indexing terms and corresponding term frequencies. [col. 4, lines 20-24]. Zhang's Figures 2A and 2B further show that both the document and document category are represented by multiple indexing terms—Indexing Term 1 through Indexing Term N—with respective Term Frequency 1 through Term Frequency N values. See [Fig. 2A and Fig. 2B] and descriptions [col. 4, lines 48 – col. 5, line 44]. Zhang's Figure 5 then expressly performs the following operations: “Document categories/archived documents with the most indexing terms similar to the document to be processed are identified. This identification may result in zero to many document categories/archived documents” [col. 8, lines 3-6] followed by: “calculating a document similarity, required for archiving documents and for identifying documents with similarities to a target document..” [col. 7, lines 54-56]. Accordingly, the rejection does not rely upon Zhang to generate Baker's structurally controlled plurality of queries, nor upon Baker alone to perform Zhang's document-similarity calculation. Baker teaches generating multiple queries according to query templates/generative rules and obtaining search-related results/scores for those queries, while Zhang teaches using multiple document-derived comparison components to calculate quantitative document-similarity values. The combined teachings therefore support the limitation for the reasons set forth in the rejection.
Claim Rejections - 35 USC § 103
14. 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.
15. 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.
16. 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.
17. Claims 1-20 are rejected under 35 U.S.C. § 103 as being unpatentable over Baker et al. US9436747B1 (hereinafter Baker) in view of Zhang et al. U.S. Patent 10,489,466 B1 (hereinafter Zhang).
Regarding claim 1, Baker discloses a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving a first document having a first schema (Baker [col. 15, lines 39-56] e.g., “The system receives (602) a set of historical queries associated with a structured document. The structured document includes embedded coding. The embedded coding describes various structures within the structured document. The structured document can include, for example, an HTML document hosted on a website. The embedded coding of the structured document can include HTML tags”. The document is associated with embedded coding such as HTML tags, which defines the schema of the document); accessing a configuration for the first schema, wherein the configuration defines how to generate, from the first document, a plurality of queries, into a collection of documents having a second schema (Baker [col. 12, lines 65 – col. 13, line 12] e.g., “The template generation module 508 generates one or more candidate query templates from the seed queries and the structured documents in the retrieved query-document pairs. To generate the candidate query templates, the template generation module 508 applies a set of template generation rules 510 to the seed queries and the structured documents.”. The system uses these templates to generate queries for a collection of documents, which implicitly involves accessing configuration the define how queries should be generated based on document schema); generating the plurality of a queries based on the configuration (Baker [col. 16, lines 22-32] e.g., “The system generates (608) the candidate synthetic queries using the query templates and other structured documents hosted on the website. Generating the candidate synthetic queries can include applying the generative rules in the query templates to the other structured documents generate the candidate synthetic queries”. These queries are derived from seed queries and structured documents, which directly corresponds to the clime limitation of generating a plurality of queries based on a configuration). Baker does not explicitly teach combining results of the plurality of queries into similarity scores for the first document. Zhang discloses combining results of the plurality of queries into similarity scores for the first document (Zhang [col. 8, lines 13-33] e.g., “…a similarity value is calculated…cosine similarity is subsequently calculated”. (Step 502), showing calculation of a similarity score for each retrieved document (Figure 5, element 502). Zhang teaches calculating similarity scores for a target document based on the results of search queries. See also [col. 8, lines 34-47] e.g., “…a determination is made about whether the similarity value is above a strong similarity threshold…”, see also [col. 7, lines 43-53] e.g., “…the most similar documents are returned … based on the ranking, may be returned”. (Figure 5, steps 504,508,510) showing aggregation and ranking to produce a similarity score for the first document. See also [col. 9, lines 41-49] e.g., “Similarity of a target document with one or more archived documents is assessed based on a quantitative measure for similarity, such as the cosine similarity”. The similarity score is then used to assess how similar the retrieved documents are to the original document. Thus, Zhang provides the explicit teaching of combining the results (features/terms/fields obtained from those searches) into a similarity score for the first document, i.e., a numeric score that expresses how similar each retrieved document is to the initial input). It would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to have combined the method and system for document similarity analysis based on weak transitive relation of similarity taught by Zhang, with query generation using structural similarity between documents taught by Baker. This combination would yield predictable and beneficial results, such as providing an enhanced ranking of documents based on their similarity to the original document.
Claims 19 and 20 incorporate substantively all the limitations of claim 1 in the form of system and method and are rejected under the same rationale.
Regarding claim 2, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the first schema is different form the second schema (Baker [col. 2, lines 21-35] e.g., “Embedding coding of the structured document can include hypertext markup language (HTML) tags”, see also (Zhang [col. 5, lines 26-38] e.g., “The document category term and term frequency list (240) in accordance with an embodiment of the technology, is equivalent to the document term and term frequency list (236), however, while the document term and term frequency list (236) is associated with a particular document (234), the document category term and term frequency list (240) is associated with a document category (238), which may include multiple or many documents (234)”. Baker’s “first schema” (structured HTML with tags) and Zhang’s “second schema” (category representation distinct from a single document). The motivation for the proposed combination is maintained.
Regarding claim 3, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the first schema defines service requests, and the second schema defines text documents comprising solutions for the service requests (Zhang [col. 7, lines43—53] e.g., “In Step 408, the most similar documents are returned, e.g., to the user who made the search request….For each of the documents in the similar document categories, a similarity value may be calculated….The document similarity value may be used to generate a ranking of the archived documents…The most similar documents, based on the ranking, may be returned”. This clearly teaches a ‘first document’ (target document) used in a search request A collection of archived documents grouped into document categories, i.e., a knowledge -base-like repository. Returning most similar documents based on a similarity score and ranking. See also (Baker [col. 4, lines 61-66] e.g., “The publishers 106a and 106b include general content servers that receive requests for content (e.g., webpages or documents related to articles, discussion threads, music, video, graphics, other webpage listings, information feeds, product reviews, etc.), and retrieve the requested content in response to the request.”. This shows, generic content servers that receive requests and return documents. The documents are text-like content (articles, discussion threads, etc.), i.e., solution-style documents served over a network. The motivation for the proposed combination is maintained.
Regarding claim 4, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration defines how to generate, from the document, queries into a plurality of collections of documents having different schemas (Baker [col. 16, lines 22-32] e.g., “The system generates (608) the candidate synthetic queries using the query templates and other structured documents hosted on the website. Generating the candidate synthetic queries can include applying the generative rules in the query templates to the other structured documents generate the candidate synthetic queries”. These queries are derived from seed queries and structured documents, which directly corresponds to the clime limitation of generating a plurality of queries based on a configuration).
Regarding claim 5, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the plurality of queries are submitted to a search interface that comprises an inverted index that accepts Boolean and phrase queries, and an Application Programming Interface (API) that receives a word and returns a number of documents in the collection of documents in which that word is used (Zhang [col. 8, lines 13-33] e.g., “…a similarity value is calculated…cosine similarity is subsequently calculated”. (Step 502), showing calculation of a similarity score for each retrieved document (Figure 5, element 502). See also [col. 8, lines 34-47] e.g., “…a determination is made about whether the similarity value is above a strong similarity threshold…”, see also [col. 7, lines 43-53] e.g., “…the most similar documents are returned … based on the ranking, may be returned”. (Figure 5, steps 504,508,510) showing aggregation and ranking to produce a similarity score for the first document. See also [col. 9, lines 41-49] e.g., “Similarity of a target document with one or more archived documents is assessed based on a quantitative measure for similarity, such as the cosine similarity”). Combining results of multiple queries into an overall similarity score is standard in document retrieval systems. The motivation for the proposed combination is maintained.
Regarding claim 6, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the first schema defines a plurality of field-value pairs (Baker [col. 2, lines 21-35] e.g., “Embedding coding of the structured document can include hypertext markup language (HTML) tags. Identifying the embedded coding fragments from the structured document can include identifying an HTML tag pair. Identifying the embedded coding fragments from the structured document can include identifying at least a portion of content enclosed by the HTML tag pair;”. The structure can have various forms, including hierarchical forms in which tags are nested. Each tag (e.g., <title>…</title>, <h1?...</h1>) acts as a field, and the text between the tags is the field’s value. This is directly a plurality of field-value pars as claimed).
Regarding claim 7, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration comprises, for a first field in the first document, a query type defining an n-gram level for a first subset of the plurality of queries (Baker [col. 10, lines 59-67] e.g., “A wildcard has the form of <type> or <type:constraint>. The “type” portion can indicate a category of terms. A “generic type” type indicates a most general type that represents either a unigram or a known n-gram. The n-gram (e.g., “new york” or “jd salinger”) can be determined by an external process”. This explains ‘query type’ (wildcard type) specifying unigram vs n-gram).
Regarding claim 8, the rejection of claim 7 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration further comprises, for the query type, a number of queries N to be generated for the query type (Baker [col. 2, lines 31-42] e.g., “Generating the query templates can include: for at least one of the embedded coding fragments, counting a number of other structured documents hosted on the website that include the embedded coding fragment; and generating the query template… in response to determining that the number of other structured documents hosted on the website including the embedded coding fragment satisfies a template qualification value”, see also [col. 3, lines 1-13] e.g., “Aggregating the query templates can include determining a template threshold, the template threshold proportional to a total number of structured documents… scoring each query template based on a number of occurrences… and aggregating the query templates that have scores satisfying the template threshold”. These passages support selecting a bounded set of templates/queries; N is an obvious explicit parameterization).
Regarding claim 9, the rejection of claim 8 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein generating the number of queries N for the query type comprises:
determining a frequency score from the collection of documents for words in the first field (Baker [col. 2, 43-67] e.g., “ The context within which the terms appear in the structured document can be based on various parameters. The parameters can include, for example, a count of a number of times the terms appear in the structured document… inverse document frequency (IDF) of the terms, among others”;
identifying the words in the first field having the N highest frequency scores (Baker [col. 4, lines 13-30] e.g., “The evaluation of the candidate synthetic queries can be measured, for example, using information retrieval (IR) scores in relation to one or more documents”; and
generating N queries from the words in the first field having the N highest frequency scores (Baker [col.2, lines 7-18] e.g., “…generating the one or more candidate synthetic queries using the query templates and other structured documents hosted on the website; measuring performance in a search operation of each of the one or more candidate synthetic queries…”. These passages directly match a frequency score based on (1) TF in the first document and (2) inverse document frequency across document).
Regarding claim 10, the rejection of claim 9 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the frequency score is determined based on a number of times a word appears in the first document and a number of documents in the collection of documents in which the word appears (Baker [col. 2, lines 43-67] e.g., “The parameters can include, for example, a count of a number of times the terms appear in the structured document, surrounding HTML tags; words or phrases that are close to one another, inverse document frequency (IDF) of the terms, among others. Generating the candidate synthetic queries can include: applying the query templates to the other structured documents (e.g., structured documents hosted on the website or a collection of websites)”).
Regarding claim 11, the rejection of claim 7 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration further comprises, for the query type, one or more target fields in the second schema (Baker [col. 2, lines 43-67] e.g., “ Generating the candidate synthetic queries can include: applying the query templates to the other structured documents…identifying other embedded coding fragments of the other structured documents that match the embedded coding fragments identified in the query templates; and designating content in the other embedded coding fragments as the candidate synthetic queries”. Here templates (config) defined for one structure/field are used against corresponding ‘embedded coding fragments’ (fields) in other document. That’s equivalent to ‘target fields’ in the second schema).
Regarding claim 12, the rejection of claim 11 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration further comprises, for a first target field in the one or more target fields, a weight to be applied to similarity scores for queries generated from the first target field (Zhang [col. 5, lines 36-44] e.g., “The document category term and term frequency list… is equivalent to the document term and term frequency list (236), however, … is associated with a document category (238)… a number of documents (234) may be assigned to a document category (238) … Similarity may be assessed… based on the indexing terms (256) … For example, a comparison of the document category… may result in a determination that the document is similar (or is not similar) to the document category”, see also [col. 8, lines 34-47] e.g., “In Step 504, a determination is made about whether the similarity value is above a strong similarity threshold… In one embodiment … a cosine similarity at or above 0.78 is considered a strong similarity….” See also [col. 8, lines 50-65] e.g., “In Step 508, a determination is made about whether the similarity value is above a weak similarity threshold… a cosine similarity of 0.2 is considered the minimum value that indicates a weak similarity. If the similarity is above the weak similarity threshold”. These show a scalar similarity value derived from weighted term contributions, and tunable thresholds. Different thresholds may be used for the detection of weak and strong similarities).
Regarding claim 13, the rejection of claim 12 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the weight is set in the configuration by a machine-learning model (Zhang [col.9, lines 1-5] e.g., “Those skilled in the art will appreciate that different thresholds may be used for the detection of weak and strong similarities. Further these thresholds may differ depending on whether the method of FIG. 5 is used for archiving a document (FIG. 3) or for performing a similarity detection (FIG. 4).” Since similarity parameters are explicitly acknowledged as tunable, learning them with a ML model trained on historical document-category relationships is an obvious optimization step)g.
Regarding claim 14, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the configuration is one of a plurality of configurations, and the plurality of configurations correspond to a plurality of different schemas (Baker [col. 2, lines 43-53] e.g., “The query template can include at least one of a literal and a wildcard, the literal including a literal phrase found in the structured document and the wildcard including a type and at least one constraint… can be based on various parameters The context …the query can include a reference to an external query configuration file that includes information about dataset sources…”, see also [col. 2, lines 54-67] e.g., “ Generating the candidate synthetic queries can include: applying the query templates to the other structured documents (e.g., structured documents hosted on the website or a collection of websites);”. Multiple sites/collections imply multiple structured conventions and therefore multiple template sets/configuration).
Regarding claim 15, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the first document is received as part of a search request to identify documents in the collection of documents that are similar to the first document (Zhang [col. 5, lines 45-52] e.g., “FIG. 4 describes the identification of documents that are similar to a target document. FIG. 5 describes the calculation of document similarity, required for archiving documents and for identifying documents with similarities to the target document”, see also, [col. 8, lines 13-33] e.g., “…a similarity value is calculated…cosine similarity is subsequently calculated”. (Step 502), showing calculation of a similarity score for each retrieved document (Figure 5, element 502), These passages support first document is received and used to identify similar documents).
Regarding claim 16, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the operations further comprise executing the plurality of queries on the collection of documents (Baker [col. 2, lines 43-67] e.g., “ Generating the candidate synthetic queries can include: applying the query templates to the other structured documents… and designating content … as the candidate synthetic queries”, see also [col. 5, lines 45-55] e.g., “The synthetic query subsystem 114 uses the synthetic query 118 to perform an augmented search operation for the query 109… The results of the augmented search operation can be provided to the user device”. This executing the (plurality) synthetic queries on a document corpus).
Regarding claim 17, the rejection of claim 16 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the results of the plurality of queries comprise scores for a second document in the collection of documents, and the scores for the second document are generated in response to the plurality of queries (Baker [col. 2, 63-67] e.g., “Measuring the performance of each of the candidate synthetic queries comprises computing a score of the candidate synthetic query in relation to the structured document. The score can be an information retrieval (IR) score.”, see also (Zhang [col. 8, lines 34-47] e.g., “In Step 504, a determination is made about whether the similarity value is above a strong similarity threshold… In one embodiment… a cosine similarity at or above 0.78 is considered a strong similarity”. These are scores for document vs. query or document vs. category). The motivation for the proposed combination is maintained.
Regarding claim 18, the rejection of claim 1 is hereby incorporated by reference, Baker and Zhang discloses a non-transitory computer-readable medium, wherein the combining the results of the plurality of queries into similarity scores comprises generating a weighted combination of scores for the second document (Zhang [col. 8, lines 14-26] e.g., “… a similarity value is calculated… The cosine similarity is subsequently calculated…”, see also [col. 9, lines 14-40] e.g., “…for document A, both category X and category Y are identified in Step 306, and accordingly document A is registered in categories X and Y. Next, when executing the method for document B, a new category Z is generated in Step 310, because no category with a strong similarity to document B was found. Document B is registered in category Z, but it is also registered in categories X and Y. Because documents A and B are registered in the same categories (X, Y), there is a substantially lower likelihood…”. Scores from multiple queries and features are combined into a single similarity value per document (cosine), which is weighted combination).
Conclusion
18. THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
19. Any inquiry concerning this communication or earlier communications from the examiner should be directed to BERHANU MITIKU whose telephone number is (571)270-1983. The examiner can normally be reached Monday – Friday 8:30AM – 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, Ajay Bhatia can be reached at 571-272-3906. 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.
/BERHANU MITIKU/Examiner, Art Unit 2156
/AJAY M BHATIA/Supervisory Patent Examiner, Art Unit 2156
Prosecution Insights