DETAILED ACTION
This action is in response to the amendment filed on 06/30/2026.
Response to Amendment
Applicant’s amendment filed on 06/30/2026 has been entered. Claims 1, 11 and 18 have been amended. No claims have been canceled. No claims have been added. Claims 1 – 20 are still pending in this application, with claims 1, 11 and 18 being independent.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 3, 5, 8 - 11, 12, 14, 16 - 18 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Apostolova (US 2024/0143642) in view of Somani et al. (US 2021/0248149) (“Somani”), and further in view of Tanaka et al. (US 2024/0211484) (“Tanaka”) and further in view of Hirsch et al. (US 11,720,580) (“Hirsch”).
For claim 1, Apostolova discloses a computer-implemented method (Abstract; [0143]) comprising: receiving, by one or more processors (Fig.6, 604; [0122] [0123] [0144]), one or more documents (first document) (Fig.5B, 505B) from a plurality of data sources (client matching system of document matching system, Fig.1,116 and 120; [0040] [0041]) ([0122] [0125]); comparing, by one or more processors (Fig.6, 604; [0122] [0123] [0144]) utilizing a fuzzy matching algorithm ([0024] [0049] [0076] [0077]), the text data from the document to a plurality of reference datasets (A reference dataset comprises a data structure which includes many data fields with corresponding values identify a project, entity and/or other organizational concept. For example a data structure corresponding to a particular delivery comprises data fields including pickup address, pickup date, pickup time, delivery address, expected delivery date, expected delivery time, actual delivery date, actual delivery time, recipient information, weight of load, driver name, and/or other data fields relevant to the data structure., [0027]) (The document matching system may identify one or more matching character strings from a document by searching the document using a plurality of character strings corresponding to a plurality of respective data fields extracted from a data structure stored on a database., Fig.4B, 410B, 415B and Fig.5B, 510B; [0093] [0100 - 0103] [0126]) to determine one or more matches between the text data and at least one of the plurality of reference datasets (“The document matching system compares the character strings to determine whether there are matching character strings between a document and a data structure. This comparison may account for near-matches or fuzzy matching dependent on the type of data field corresponding to the particular character strings being compared.”, [0049] [0075] [0076] [0103] [0126]), wherein the one or more matches are based on at least one similarity score (match threshold/score, wherein the match threshold may be 80% matching characters, [0076] [0103]); inputting, by one or more processors (Fig.6, 604; [0122] [0123] [0144]), the determined one or more matches into a trained machine-learning model to refine the one or more matches (Fig.4B, 420B and Fig.5B, 515B; [0104] [0127]); and outputting, by the one or more processors, a representation of the refined one or more matches to a graphical user interface of a device (Fig.4B, 425B and Fig.5B, 520B; [0105] [0128]).
Yet, Apostolova fails to teach the following: extracting, by the one or more processors utilizing an optical character recognition algorithm, text data from the one or more documents and comparing the extracted text data to the reference dataset; inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches, wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon the extracted text data and the plurality of reference datasets; and outputting the one or more matches with the at least one similarity score.
However, a separate embodiment of Apostolova discloses the following: utilizing an optical character recognition algorithm to extract character string data from a document ([0074]); and comparing the extracted character string data to determine matches ([0075]).
Moreover, Somani discloses a system and method for performing entity matching (Abstract), comprising the following: generating a set of first similarity scores ([0027 – 0038]) between attributes of matching data records ([0044] [0045] [0060]); and inputting the similarity scores and matching attributes from different data records into a trained machine learning model to execute a model scoring to determine final similarity scores which refine the matches ([0014] [0045] [0061]).
Additionally, Tanaka discloses an information processing method and apparatus (Abstract), comprising the following: a similar record/entity match method is performed ([0095 – 0101]); and a calculated similarity score is displayed along with the matches (Fig.9, 302; [0101 – 0106]).
Furthermore, Hirsch discloses a fuzzy matching system (Abstract), comprising the following for refining one or more matches: detecting, by one or more processors (column 2 lines 49 – 56; claim 18) utilizing a trained machine-learning model (Fig.1, 130, column 5 lines 59 – column 6 line 6), one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score (User provides feedback on matches with at least 5% probability/similarity score generated by a fuzzy algorithm. The user feedback indicates discrepancies., Fig.4, 440, 442 and 446, Fig.6, Fig.7; column 5 lines 39 – 58; column 6 lines 36 – 51; column 7 line 31 – column 8 line 13; column 8 line 38 – column 9 line7; claim 1); adjusting, by the one or more processors utilizing a trained machine-learning model one or more parameters (parameters/weights) of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies (column 1 lines 37 – 44; column 9 lines 4 – 7;claim 4); and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, one or more matches based upon pairs of data (An entity resolution process with the updated fuzzy matching algorithm is run in the future., column 6 lines 8 – 19).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to modify Apostolova’s first embodiment with Apostolova’s second embodiment so that the following occurs for the purpose of improving the automation of document review and management by accounting for documents which are images/scans which have not been pre-processed ([0054] [0060]): further utilizing an optical character recognition to extract character string data from the one or more documents; and further comparing the extracted string data to determine the matches.
Moreover, it would have been obvious to one of ordinary skull in the art at the time of applicant’s filing to improve Apostolova’s invention in the same way that Somani’s invention has been improved to achieve the following, predictable results for the purpose of reducing the effects of data misidentifications while performing entity matching with data records (Somani, [0010 – 0013]): further training the machine learning model to process both matches and the similarity scores; and further inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches.
Additionally, it would have been obvious to one of ordinary skill in the art the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova and Somani in the same way that Tanaka’s invention has been improved to achieve the following, predictable results for the purpose of increasing user satisfaction while performing database entity matching (Tanaka, [0003] [0004]): further outputting the one or more matches with the at least one similarity score.
Furthermore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova, Somani and Tanaka in the same way that Hirsch’s invention has been improved to achieve the following, predictable results for the purpose of improving candidate pair suggestions in entity matching by decreasing false positive matches and increasing true positive matches (Hirsch, column 6 lines 8 – 20): wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon data pairs, e.g. extracted text data and the plurality of reference datasets.
For claim 11, Apostolova discloses a system (Abstract) comprising: one or more processors of a computing system (Fig.6, 604; [0122] [0123] [0143] [0144]); and at least non-transitory computer readable medium (Abstract; Fig.6, 608; [0122] [0123] [0143] [0147] [0155]), the non-transitory computer readable medium storing instructions ([0147] [0155]) which, when executed by one or more processors of a computing system, cause the one or more processors to perform operations ([0155]) comprising: receiving one or more documents (first document, Fig.5B, 505B) from a plurality of data sources (client matching system of document matching system, Fig.1,116 and 120; [0040] [0041]) ([0122] [0125]); comparing, utilizing a matching algorithm (fuzzy matching algorithm), text data to a plurality of reference datasets (A reference dataset comprises a data structure which includes many data fields with corresponding values identify a project, entity and/or other organizational concept. For example a data structure corresponding to a particular delivery comprises data fields including pickup address, pickup date, pickup time, delivery address, expected delivery date, expected delivery time, actual delivery date, actual delivery time, recipient information, weight of load, driver name, and/or other data fields relevant to the data structure., [0027]) (The document matching system may identify one or more matching character strings from a document by searching the document using a plurality of character strings corresponding to a plurality of respective data fields extracted from a data structure stored on a database., Fig.4B, 410B, 415B and Fig.5B, 510B; [0093] [0100 - 0103] [0126]) to determine one or more matches between the text data and at least one of the plurality of reference datasets (“The document matching system compares the character strings to determine whether there are matching character strings between a document and a data structure. This comparison may account for near-matches or fuzzy matching dependent on the type of data field corresponding to the particular character strings being compared.”, [0049] [0075] [0076] [0103] [0126]), wherein the one or more matches are based on at least one similarity score (match threshold/score, wherein the match threshold may be 80% matching characters, [0076] [0103]); inputting the determined one or more matches into a trained machine-learning model to validate a similarity assessment (Fig.4B, 420B and Fig.5B, 515B; [0104] [0127]); and outputting a representation of the one or more matches to a graphical user interface of a device (Fig.4B, 425B and Fig.5B, 520B; [0105] [0128]).
Yet, Apostolova fails to teach the following: extracting, utilizing an extraction technology, text data from the one or more documents and comparing the extracted text data to the reference dataset; inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches, wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon the extracted text data and the plurality of reference datasets; and outputting the one or more matches with the at least one similarity score; and outputting the one or more matches with the at least one similarity score.
However, a separate embodiment of Apostolova discloses the following: utilizing an optical character recognition algorithm to extract character string data from a document ([0074]); and comparing the extracted character string data to determine matches ([0075]).
Moreover, Somani discloses a system and method for performing entity matching (Abstract), comprising the following: generating a set of first similarity scores [0027 – 0038]) between attributes of matching data records ([0044] [0045]); and inputting the similarity scores and matching attributes from different data records into a trained machine learning model to execute a model scoring to determine final similarity scores which refine the matches ([0014]) [0045]).
Additionally, Tanaka discloses an information processing method and apparatus (Abstract), comprising the following: a similar record/entity match method is performed ([0095 – 0101]); and a calculated similarity score is displayed along with the matches (Fig.9, 302; [0101 – 0106]).
Furthermore, Hirsch discloses a fuzzy matching system (Abstract), comprising the following for refining one or more matches: detecting, by one or more processors (column 2 lines 49 – 56; claim 18) utilizing a trained machine-learning model (Fig.1, 130, column 5 lines 59 – column 6 line 6), one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score (User provides feedback on matches with at least 5% probability/similarity score generated by a fuzzy algorithm. The user feedback indicates discrepancies., Fig.4, 440, 442 and 446, Fig.6, Fig.7; column 5 lines 39 – 58; column 6 lines 36 – 51; column 7 line 31 – column 8 line 13; column 8 line 38 – column 9 line7; claim 1); adjusting, by the one or more processors utilizing a trained machine-learning model one or more parameters (parameters/weights) of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies (column 1 lines 37 – 44; column 9 lines 4 – 7;claim 4); and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, one or more matches based upon pairs of data (An entity resolution process with the updated fuzzy matching algorithm is run in the future., column 6 lines 8 – 19)
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to modify Apostolova’s first embodiment with Apostolova’s second embodiment so that the following occurs for the purpose of improving the automation of document review and management by accounting for documents which are images/scans which have not been pre-processed ([0054] [0060]): further utilizing an optical character recognition to extract character string data from the one or more documents; and further comparing the extracted string data to determine the matches.
Moreover, it would have been obvious to one of ordinary skull in the art at the time of applicant’s filing to improve Apostolova’s invention in the same way that Somani’s invention has been improved to achieve the following, predictable results for the purpose of reducing the effects of data misidentifications while performing entity matching with data records (Somani, [0010 – 0013]): further training the machine learning model to process both matches and the similarity scores; and further inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches.
Additionally, it would have been obvious to one of ordinary skill in the art the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova and Somani in the same way that Tanaka’s invention has been improved to achieve the following, predictable results for the purpose of increasing user satisfaction while performing database entity matching (Tanaka, [0003] [0004]): further outputting the one or more matches with the at least one similarity score.
Furthermore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolva, Somani and Tanaka in the same way Hirsch’s invention has been improved to achieve the following, predictable results for the purpose of improving candidate pair suggestions in entity matching by decreasing false positive matches and increasing true positive matches (Hirsch, column 6 lines 8 – 20): wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon data pairs, e.g. extracted text data and the plurality of reference datasets.
For claim 18, Apostolova discloses a non-transitory computer readable medium (Abstract; Fig.6, 608; [0122] [0123] [0143] [0147] [0155]), the non-transitory computer readable medium storing instructions ([0147] [0155]) which, when executed by one or more processors of a computing system, cause the one or more processors to perform operations ([0155]) comprising: receiving, by one or more processors (Fig.6, 604; [0122] [0123] [0144]), one or more documents (first document, Fig.5B, 505B) from a plurality of data sources (client matching system of document matching system, Fig.1,116 and 120; [0040] [0041]) ([0122] [0125]); comparing, by one or more processors (Fig.6, 604; [0122] [0123] [0144]) utilizing a fuzzy matching algorithm ([0024] [0049] [0076] [0077]), the text data from the document to a plurality of reference datasets (A reference dataset comprises a data structure which includes many data fields with corresponding values identify a project, entity and/or other organizational concept. For example a data structure corresponding to a particular delivery comprises data fields including pickup address, pickup date, pickup time, delivery address, expected delivery date, expected delivery time, actual delivery date, actual delivery time, recipient information, weight of load, driver name, and/or other data fields relevant to the data structure., [0027]) (The document matching system may identify one or more matching character strings from a document by searching the document using a plurality of character strings corresponding to a plurality of respective data fields extracted from a data structure stored on a database., Fig.4B, 410B, 415B and Fig.5B, 510B; [0093] [0100 - 0103] [0126]) to determine one or more matches between the text data and at least one of the plurality of reference datasets (“The document matching system compares the character strings to determine whether there are matching character strings between a document and a data structure. This comparison may account for near-matches or fuzzy matching dependent on the type of data field corresponding to the particular character strings being compared.”, [0049] [0075] [0076] [0103] [0126]), wherein the one or more matches are based on at least one similarity score (match threshold/score, wherein the match threshold may be 80% matching characters, [0076] [0103]); inputting, by one or more processors (Fig.6, 604; [0122] [0123] [0144]), the determined one or more matches into a trained machine-learning model to refine the one or more matches (Fig.4B, 420B and Fig.5B, 515B; [0104] [0127]); and outputting, by the one or more processors, a representation of the refined one or more matches to a graphical user interface of a device (Fig.4B, 425B and Fig.5B, 520B;[0105] [0128]).
Yet, the first embodiment Apostolova fails to teach the following: extracting, by the one or more processors utilizing an optical character recognition algorithm, text data from the one or more documents and comparing the extracted text data to the reference dataset; inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches, wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon the extracted text data and the plurality of reference datasets; and outputting the one or more matches with the at least one similarity score; and outputting the one or more matches with the at least one similarity score.
However, a second embodiment of Apostolova discloses the following: utilizing an optical character recognition algorithm to extract character string data from a document ([0074]); and comparing the extracted character string data to determine matches ([0075]).
Moreover, Somani discloses a system and method for performing entity matching (Abstract), comprising the following: generating a set of first similarity scores( [0027 – 0038]) between attributes of matching data records ([0044] [0045]); and inputting the similarity scores and matching attributes from different data records into a trained machine learning model to execute a model scoring to determine final similarity scores which refine the matches ([0014] [0045]).
Additionally, Tanaka discloses an information processing method and apparatus (Abstract), comprising the following: a similar record/entity match method is performed ([0095 – 0101]); and a calculated similarity score is displayed along with the matches (Fig.9, 302; [0101 – 0106]).
Furthermore, Hirsch discloses a fuzzy matching system (Abstract), comprising the following for refining one or more matches: detecting, by one or more processors (column 2 lines 49 – 56; claim 18) utilizing a trained machine-learning model (Fig.1, 130, column 5 lines 59 – column 6 line 6), one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score (User provides feedback on matches with at least 5% probability/similarity score generated by a fuzzy algorithm. The user feedback indicates discrepancies., Fig.4, 440, 442 and 446, Fig.6, Fig.7; column 5 lines 39 – 58; column 6 lines 36 – 51; column 7 line 31 – column 8 line 13; column 8 line 38 – column 9 line7; claim 1); adjusting, by the one or more processors utilizing a trained machine-learning model one or more parameters (parameters/weights) of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies (column 1 lines 37 – 44; column 9 lines 4 – 7;claim 4); and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, one or more matches based upon pairs of data (An entity resolution process with the updated fuzzy matching algorithm is run in the future., column 6 lines 8 – 19)
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to modify Apostolova’s first embodiment with Apostolova’s second embodiment so that the following occurs for the purpose of improving the automation of document review and management by accounting for documents which are images which have not been pre-processed ([0054] [0060]): further utilizing an optical character recognition to extract character string data from the one or more documents; and further comparing the extracted string data to determine the matches.
Moreover, it would have been obvious to one of ordinary skull in the art at the time of applicant’s filing to improve Apostolova’s invention in the same way that Somani’s invention has been improved to achieve the following, predictable results for the purpose of reducing the effects of data misidentifications while performing entity matching with data records (Somani, [0010 – 0013]): further training the machine learning model to process both matches and the similarity scores; and further inputting the determined one or more matches and the at least one similarity score into the trained machine learning model to refine the one or more matches.
Additionally, it would have been obvious to one of ordinary skill in the art the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova and Somani in the same way that Tanaka’s invention has been improved to achieve the following, predictable results for the purpose of increasing user satisfaction while performing database entity matching (Tanaka, [0003] [0004]): further outputting the one or more matches with the at least one similarity score.
Furthermore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolva, Somani and Tanaka in the same way that Hirsch’s invention has been improved to achieve the following, predictable results for the purpose of improving candidate pair suggestions in entity matching by decreasing false positive matches and increasing true positive matches (Hirsch, column 6 lines 8 – 20): wherein refining the one or more matches includes: detecting, by the one or more processors utilizing the trained machine-learning model, one or more patterns or one or more discrepancies corresponding to the determined one or more matches based on the at least one similarity score; adjusting, by the one or more processors utilizing the trained machine-learning model, one or more parameters of the fuzzy matching algorithm based on the detected one or more patterns or one or more discrepancies; and updating, utilizing the fuzzy matching algorithm with the adjusted one or more parameters, the one or more matches based upon data pairs, e.g. extracted text data and the plurality of reference datasets.
For claims 3 and 20, Apostolova further discloses, wherein comparing the extracted text data to the plurality of reference datasets for determining the one or more matches comprises: calculating, by the one or more processors utilizing the fuzzy matching algorithm, the at least one similarity score for the extracted text data based on one or more factors, wherein the one or more factors include an edit distance, a token-based similarity algorithm (Apostolova, [0074 – 0076] [0103] [0126]), or a contextual relevance; and determining, by the one or more processors utilizing the fuzzy matching algorithm, the one or more matches by evaluating the at least one similarity score against a pre-determined threshold (Apostolova, [0076] [0103] [0126]) wherein the pre-determined threshold indicates a minimum acceptable similarity level for the one or more matches (Apostolova, [0076] [0103] [0126]).
For claim 5, Apostolova further discloses, wherein the token-based similarity algorithm measures a degree of similarity between extracted text data and at least one of the plurality of reference datasets (Apostolova, [0075] [0076] [0103] [0126]), and wherein the degree of similarity includes one or more common substrings (Apostolova, [0075] [0076] [0103] [0126]) or a phonetic resemblance.
For claim 8, Apostolova further discloses, wherein the fuzzy matching algorithm performs partial matching by identifying and scoring individual segments of the extracted text data against the plurality of reference datasets (Apostolova, [0024] [0074– 0077] [0103] [0126]).
For claims 9 and 16, the combination of Apostolova’s and Hirsch further disclose wherein the fuzzy matching algorithm utilizes one or more similarity metrics to compare the extracted text data to the plurality of reference datasets (Apostolova, [0075] [0076] [0103] [0126]), and wherein the one or more similarity metrics include a Levenshtein distance (Hirsch,Fig.3A, 330; column 6 lines 55 – column 7 line 19) or a Jaccard similarity.
For claims 10 and 17, Apostolova and Hirsch further disclose: wherein the fuzzy matching algorithm utilizes one or more phonetic algorithms for handling one or more variations in spelling or pronunciations of the extracted text data and the plurality of reference datasets, and wherein the one or more phonetic algorithms include a Soundex algorithm (Apostolova, [0049] [0075] [0076] [0103] [0126]) (Hirsch, Fig.3A, 330; column 6 lines 55 – column 7 line 19) or a Metaphone algorithm
For claim 12, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch further discloses, wherein inputting the determined one or more matches and the at least one similarity score into the trained machine-learning model to validate the similarity assessment comprises: analyzing, utilizing the trained machine-learning model, the determined one or more matches and the at least one similarity score (Apostolova, [0104]) (Somani, [0014] [0044] [0045]).
Yet, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach adjusting one or more parameters of the similarity assessment based on the analysis.
However, a third embodiment of Apostolova discloses the use of feedback to adjust (re-train) one or more parameters (first machine learning process) of a similarity assessment (Apostolova, [0057] [0058] [0061] [0062] [0065] [0067] [0068] [0070]).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch in the same way that Apostolova’s third embodiment has been improved to achieve the following, predictable results for the purpose of improving a matching accuracy when performing document matching between all types of documents including poor quality versions of documents (Apostolova, [0003 – 0005]): further providing a feedback loop to receive feedback based on the analysis (output of the trained machine learning model) of the determined one or more matches and the at least one similarity score; and further adjusting one or more parameters (Apostolova, first machine learning process, [0101]) of the similarity assessment.
For claim 14, Apostolova further discloses, wherein the matching algorithm includes a fuzzing matching algorithm (Apostolova, [0024] [0049] [0076] [0077]), and wherein comparing the extracted text data to the plurality of reference datasets for determining the one or more matches comprises: calculating, by the one or more processors utilizing the fuzzy matching algorithm, the at least one similarity score for the extracted text data based on one or more factors, wherein the one or more factors include an edit distance, a token-based similarity algorithm (Apostolova, [0074 – 0076] [0103] [0126]), or a contextual relevance; and determining, utilizing the fuzzy matching algorithm, the one or more matches by evaluating the at least one similarity score against a pre-determined threshold (Apostolova, [0076] [0103] [0126]) wherein the pre-determined threshold indicates a minimum acceptable similarity level for the one or more matches (Apostolova, [0076] [0103] [0126]).
Claim(s) 2, 13 and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Apostolova (US 2024/0143642) in view of Somani et al. (US 2021/0248149) (“Somani”), and further in view of Tanaka et al. (US 2024/0211484) (“Tanaka”) and further in view of Hirsch et al. (US 11,720,580) (“Hirsch”) and further in view of UbiAI (“What is OCR in 2024?”).
For claims 2 and 19, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch further discloses, wherein extracting the text data from the one or more documents comprises: processing, by the one or more processors utilizing the optical character recognition algorithm, the text data for identifying and segmenting text regions in the one or more documents (Apostolova, [0054] [0060] [0074]); recognizing, by the one or more processors utilizing the optical character recognition algorithm, characters within the segmented text regions for extraction (Apostolova, [0054] [0060] [0074]).
Yet, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach the following: generating, by the one or more processors utilizing the optical character recognition algorithm, a digital representation of the extracted text data in a machine-readable format.
However, UbiAI discloses a method for digitizing physical documents (Abstract), wherein OCR is used to generate a digital representation of extracted data in a machine readable format (Importance of OCR and The Mechanics of OCR/ 3. Text Recognition and 4. Post Processing).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch in the same way that UbiAI’s invention has been improved to achieve the following, predictable results for the purpose of enabling text-based data analysis and manipulation of paper-based documentation (UbiAI, Importance of OCR): further generating, by the one or more processors utilizing the optical character recognition algorithm, a digital representation of the extracted text data in a machine-readable format.
For claim 13, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch further discloses, wherein the extraction technology includes an optical character recognition algorithm (Apostolova, [0054] [0060] [0074]) and wherein extracting the text data from the one or more documents comprises: processing, utilizing the optical character recognition algorithm, the text data for identifying and segmenting text regions in the one or more documents (Apostolova, [0054] [0060] [0074]); and recognizing, utilizing the optical character recognition algorithm, characters within the segmented text regions for extraction(Apostolova, [0054] [0060] [0074]).
Yet, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach the following: generating, by the one or more processors utilizing the optical character recognition algorithm, a digital representation of the extracted text data in a machine-readable format.
However, UbiAI discloses a method for digitizing physical documents (Abstract), wherein OCR is used to generate a digital representation of extracted data in a machine readable format (Importance of OCR and The Mechanics of OCR/ 3. Text Recognition and 4. Post Processing).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova, Somani, Tanaka and Hirsch in the same way that UbiAI’s invention has been improved to achieve the following, predictable results for the purpose of enabling text-based data analysis and manipulation of paper-based documentation (UbiAI, Importance of OCR): further generating, by the one or more processors utilizing the optical character recognition algorithm, a digital representation of the extracted text data in a machine-readable format.
Claim(s) 4 and 6 is/are rejected under 35 U.S.C. 103 as being unpatentable over Apostolova (US 2024/0143642) in view of Somani et al. (US 2021/0248149) (“Somani”), and further in view of Tanaka et al. (US 2024/0211484) (“Tanaka”), and further in view of Hirsch et al. (US 11,720,580) (“Hirsch”) and further in Kuruvilla (“What is Fuzzy Search and Fuzzy Matching”).
For claim 4, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach the following, wherein the edit distance measures a minimum number of single-character edits for transforming the extracted text data into at least one of the plurality of reference datasets.
However, Kuruvilla discloses a method for performing fuzzy matching of texts (“What is Fuzzy Search” and “What is Fuzzy Matching”), comprising the following: a fuzzy matching algorithm used to calculate a similarity score for text data comprises edit distance (Levenshtein Distance), wherein the edit measures a minimum number of single-character edits for transforming a first text string into second text string (“How does Fuzzy Name/Fuzzy String Matching Work? and How to Perform Fuzzy Name Matching, 1) Levenshtein Distance).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch in the same way Kuruvilla’s invention has been improved to achieve the following, predictable results for the purpose of considering fuzzy or near matches when performing document matching between all types of documents including poor quality versions of documents (Apostolova, [0003 – 0005]): the fuzzy matching algorithm is further used to calculate the similarity score based on edit distance; and the edit distance measures a minimum number of single-character edits for transforming first text data, e.g. the extracted text data , into at least one of the plurality of second text data, e.g. one of the plurality of reference datasets.
For claim 6, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach the following: processing, by the one or more processors, the extracted text data by utilizing a natural language processing (NLP) algorithm; and determining, by the one or more processors, a semantic meaning or a contextual alignment between the extracted text data and the plurality of reference datasets.
However, Kuruvilla discloses a method for performing fuzzy matching of texts (“What is Fuzzy Search” and “What is Fuzzy Matching”), comprising the following: text data is processed utilizing a NLP algorithm (The names are split into corresponding “n-gram” representations; and each n-gram is vectorized by using an encoding technique., How does Fuzzy Name/Fuzzy String Matching Work? and How to Perform Fuzzy Name Matching, 4) Cosine Similarity); and a contextual alignment is generated between the text data and a second text data (A cosine similarity is computed between two vector representations of strings., How to Perform Fuzzy Name Matching, 4) Cosine Similarity).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch in the same way that Kuruvilla’s invention has been improved to achieve the following, predictable results for the purpose of considering fuzzy or near matches when performing document matching between all types of documents including poor quality versions of documents (Apostolova, [0003 – 0005]): performing fuzzy matching by further processing text data, e.g. the extracted text data, by utilizing a natural language processing (NLP) algorithm; and further determining a semantic meaning or a contextual alignment between the extracted text data and second text data, e.g. the plurality of reference datasets.
Claim(s) 7 and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Apostolova (US 2024/0143642) in view of Somani et al. (US 2021/0248149) (“Somani”), and further in view of Tanaka et al. (US 2024/0211484) (“Tanaka”), and further in view of Hirsch et al. (US 11,720,580) (“Hirsch”) and further in Singh et al. (US 2018/0308003) (“Singh”).
For claims 7 and 15, the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch fails to teach the following: wherein determining the one or more matches by evaluating the at least one similarity score against the pre-determined threshold comprises: calculating, by the one or more processors utilizing the fuzzy matching algorithm, the at least one similarity score by aggregating the one or more factors into a composite similarity score for each comparison; and selecting, by the one or more processors utilizing the fuzzy matching algorithm, the text data from the plurality of reference datasets upon determining the composite similarity score exceeds the pre-determined threshold.
However, Singh discloses a hybrid method for performing string matching (Abstract), comprising the following for comparing first text data (incomplete input string) and reference text data strings stored in a database): calculating a distance similarity metric (Levenshtein, [0036] [0098]) between the first text data and reference text data ((Fig.2, 206B; [0056 – 0060] [0116 – 0124] [0225); calculating a phonetic similarity metric (Soundex, [0035] [0098]) between the first text data and reference text data (Fig.2, 206A; [0056 – 0060] [0099 – 0104]); calculating a similarity score by aggregating the distance similarity metric and phonetic similarity metric into a composite similarity score (Fig.2, 208; [0061] [0172 – 0177]); and selecting the reference text data upon determining that the composite similarity score exceeds a pre-determined threshold (The stored strings with similarity scores above a threshold are output., Fig.2, 210; [0007] [0062]).
Therefore, it would have been obvious to one of ordinary skill in the art at the time of applicant’s filing to improve the invention disclosed by the combination of Apostolova’s embodiments, Somani, Tanaka and Hirsch in the same way that Singh’s invention has been improved to achieve the following, predictable results for the purpose of improving the accuracy of performing document matching between all types of documents including documents which may comprise incomplete/incorrect text (Apostolova, [0003 – 0005]) (Singh, [0001 – 0006]), wherein determining the one or more matches by evaluating the at least one similarity score against the pre-determined threshold comprises: calculating, by the one or more processors utilizing the fuzzy matching algorithm, a factor including an edit distance factor and a second similarity metric comprising a phonetic metric; further calculating, by the one or more processors utilizing the fuzzy matching algorithm, the at least one similarity score by aggregating the one or more factors, e.g. the edit distance, into a composite (based on both the edit distance and phonetic metric) similarity score for each comparison; and further selecting, by the one or more processors utilizing the fuzzy matching algorithm, the text data from the plurality of reference datasets upon determining the composite similarity score exceeds the pre-determined threshold.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1 – 20 have been considered but are moot in view of the new ground(s) of rejection.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SONIA L GAY whose telephone number is (571)270-1951. The examiner can normally be reached Monday-Friday 9-5 ET.
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, Daniel Washburn can be reached at 571-272-5551. 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.
/SONIA L GAY/Primary Examiner, Art Unit 2657
Prosecution Insights