# US 12,216,685 B2 · Systems and methods for pattern-based multi-stage deterministic data classification > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,216,685 B2 - Document ID: US12216685B2 - Status: Granted - Kind: B2 - Application number: 18/394,744 - Filed: 2023-12-22 - Published: 2024-04-18 - Granted: 2025-02-04 - Earliest claimed: 2022-04-15 - Assignees: The Travelers Indemnity Company - Applicants: The Travelers Indemnity Company - Inventors: Zi Cheng Feng - Web page: https://patentgrep.com/patent/US12216685B2 ## CPC - G06F 16/285: {Clustering or classification} - G06V 30/10: Character recognition - G06F 16/355: Creation or modification of classes or clusters - G06F 16/55: Clustering; Classification - G06F 16/5846: {using extracted text} - G06V 30/19: Recognition using electronic means - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract Systems and methods for pattern-based multi-stage deterministic data classification that may reduce processing and memory overhead while providing more accurate data classifications. ## Claims ### Claim 1 A method for pattern-based multi-stage deterministic data classification, comprising: receiving, by at least one electronic processing device from a plurality of electronic processing devices of a server, an application for an underwriting product, the application comprising at least one data value and at least one attached file; converting, utilizing an Optical Character Recognition (OCR) algorithm stored in a non-transitory data storage device that is in communication with the at least one electronic processing device, at least one portion of the at least one attached file into at least one searchable text character, thereby defining a searchable version of the at least one attached file; searching, by the at least one electronic processing device and utilizing a listing of keywords stored in the non-transitory data storage device, the listing of keywords representing required data elements for a consolidated set of strategic classification rules, the searchable version of the at least one attached file; identifying, by the at least one electronic processing device and based on the searching, and for each required data element for the consolidated set of strategic classification rules, whether a value for the required data element exists in the at least one attached file; categorizing, by the at least one electronic processing device and based on the identifying of whether the values for the required data elements exist in the at least one attached file, the application into one of a plurality of pre-defined data availability categories; selecting, by the at least one electronic processing device and based on the categorized one of the plurality of pre-defined data availability categories for the application and utilizing a mapping stored in the non-transitory data storage device, the mapping comprising a mapping of the plurality of data availability categories to a plurality of signature analysis routines, one of the plurality of signature analysis routines; identifying, by the at least one electronic processing device and by an execution of the selected one of the plurality of signature analysis routines, values for the required data elements for the consolidated set of strategic classification rules; evaluating, by the at least one electronic processing device and based on the identified values for the required data elements for the consolidated set of strategic classification rules, the consolidated set of strategic classification rules; identifying, by the at least one electronic processing device and based on the evaluating, a numeric value for a classification; and automatically routing, by the at least one electronic processing device, the application for the underwriting product to an address assigned to the identified classification. ### Claim 2 (depends on 1) The method of claim 1, further comprising: conducting a pattern-based multi-stage deterministic strategic classification rule consolidation, comprising: identifying a set of strategic classification rules, the set of strategic classification rules comprising a plurality of data rows; identifying, for each data row of the plurality of data rows of the set of strategic classification rules, at least one assigned condition; identifying a plurality of commonalities between assigned conditions for different data rows of the set of strategic classification rules; and grouping, based on the identified commonalities, at least two different data rows of the set of strategic classification rules, thereby defining the consolidated set of strategic classification rules. ### Claim 3 (depends on 2) The method of claim 2, wherein the identifying of the set of strategic classification rules comprises: receiving a spreadsheet file comprising a plurality of data rows defining the set of strategic classification rules. ### Claim 4 (depends on 3) The method of claim 3, wherein the set of strategic classification rules comprises a plurality of numeric threshold definitions. ### Claim 5 (depends on 3) The method of claim 3, wherein the plurality of data rows comprises a number of rows greater than one thousand. ### Claim 6 (depends on 2) The method of claim 2, wherein the identifying of the set of strategic classification rules comprises: querying a database comprising a plurality of data entries storing definitions of the set of strategic classification rules. ### Claim 7 (depends on 6) The method of claim 6, wherein the plurality of data entries comprises a number of entries greater than one thousand. ### Claim 8 (depends on 2) The method of claim 2, wherein the grouping of the at least two different data rows of the set of strategic classification rules comprises at least one of (i) consolidating the at least two different rows of the set of strategic classification rules into a single row defining the consolidated set of strategic classification rules and (ii) merging the at least two different rows of the set of strategic classification rules into the single row defining the consolidated set of strategic classification rules. ### Claim 9 (depends on 1) The method of claim 1, wherein the categorized one of the plurality of pre-defined data availability categories comprises a category wherein one of: (i) the values for all the required data elements are identified to exist in the application, (ii) a subset of the values for all the required data elements are identified to exist in the application, or (iii) no values for the required data elements are identified to exist in the application. ### Claim 10 (depends on 1) The method of claim 1, wherein the categorizing of the application into the one of the plurality of pre-defined data availability categories is further based on the at least one data value of the application. ### Claim 11 (depends on 10) The method of claim 10, wherein the at least one data value comprises a JSON value. ### Claim 12 (depends on 1) The method of claim 1, wherein the categorized one of the plurality of pre-defined data availability categories comprises a category wherein each desired data element for the application: (i) is represented by an identified corresponding keyword in the at least one attached file, (ii) is represented by the at least one data value of the application, or (iii) can be calculated based on one or more data elements from (i) or (ii). ### Claim 13 (depends on 1) The method of claim 1, wherein the categorized one of the plurality of pre-defined data availability categories comprises a category wherein only a subset of all the required data elements are identified to exist in the application. ### Claim 14 (depends on 2) The method of claim 2, wherein the consolidated set of strategic classification rules are defined, based on the grouping, by less than fifty data rows. ### Claim 15 (depends on 1) The method of claim 1, wherein the identifying of the values for the required data elements for the consolidated set of strategic classification rules comprises extracting the values from the at least one attached file of the application. ### Claim 16 (depends on 1) The method of claim 1, wherein the identifying of the values for the required data elements for the consolidated set of strategic classification rules comprises identifying the values from the at least one data value of the application. ### Claim 17 (depends on 1) The method of claim 1, wherein the identifying of the values for the required data elements for the consolidated set of strategic classification rules comprises calculating the values from one or more of (i) the at least one data value of the application and (ii) at least one value extracted from the at least one attached file of the application. ### Claim 18 A method for pattern-based multi-stage deterministic data classification, comprising: receiving, by at least one electronic processing device from a plurality of electronic processing devices of a server, an application for an underwriting product, the application comprising at least one data value and at least one attached file; converting, utilizing an Optical Character Recognition (OCR) algorithm stored in a non-transitory data storage device that is in communication with the at least one electronic processing device, at least one portion of the at least one attached file into at least one searchable text character, thereby defining a searchable version of the at least one attached file; searching, by the at least one electronic processing device and utilizing a listing of keywords stored in the non-transitory data storage device, the listing of keywords representing required data elements for a consolidated set of strategic classification rules, wherein the consolidated set of strategic classification rules are defined utilizing a pattern-based multi-stage deterministic consolidation process, the searchable version of the at least one attached file; identifying, by the at least one electronic processing device and based on the searching, and for each required data element for the consolidated set of strategic classification rules, whether a value for the required data element exists in the at least one attached file; categorizing, by the at least one electronic processing device and based on the identifying of whether the values for the required data elements exist in the at least one attached file and based on the at least one data value of the application, the application into one of a plurality of pre-defined data availability categories; selecting, by the at least one electronic processing device and based on the categorized one of the plurality of pre-defined data availability categories for the application and utilizing a mapping stored in the non-transitory data storage device, the mapping comprising a mapping of the plurality of data availability categories to a plurality of signature analysis routines, one of the plurality of signature analysis routines; identifying, by the at least one electronic processing device and by an execution of the selected one of the plurality of signature analysis routines, values for the required data elements for the consolidated set of strategic classification rules; evaluating, by the at least one electronic processing device and based on the identified values for the required data elements for the consolidated set of strategic classification rules, the consolidated set of strategic classification rules; identifying, by the at least one electronic processing device and based on the evaluating, a numeric value for a classification; and automatically routing, by the at least one electronic processing device, the application for the underwriting product to an address assigned to the identified classification. ### Claim 19 (depends on 18) The method of claim 18, wherein the at least one data value comprises a JSON value. ### Claim 20 (depends on 18) The method of claim 18, wherein the categorized one of the plurality of pre-defined data availability categories comprises a category wherein one of: (i) the values for all the required data elements are identified to exist in the application, (ii) a subset of the values for all the required data elements are identified to exist in the application, or (iii) no values for the required data elements are identified to exist in the application.