# US 11,308,143 B2 · Discrepancy curator for documents in a corpus of a cognitive computing system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,308,143 B2 - Document ID: US11308143B2 - Status: Granted - Kind: B2 - Application number: 16/375,057 - Filed: 2019-04-04 - Published: 2019-08-01 - Granted: 2022-04-19 - Earliest claimed: 2016-01-12 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Donna K. Byron; Elie Feirouz; Ashok Kumar; William G. O'Keeffe - Web page: https://patentgrep.com/patent/US11308143B2 ## CPC - G06F 16/355: Creation or modification of classes or clusters - G06F 16/335: Filtering based on additional data, e.g. user or group profiles (filtering in web context G06F16/9535, G06F16/9536) - G06F 16/338: Presentation of query results - G06F 40/226: Validation - G06N 20/00: Machine learning - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Curation of a corpus of a cognitive computing system is performed by reporting to a user a cluster model of a parse tree structure of discrepancies and corresponding assigned confidence factors detected between at least a portion of a first electronic document and a second or more electronic documents in the information corpus. Responsive to a selection by the user of a discrepancy cluster model, drill-down details regarding the discrepancy are returned to the user, for subsequent user selection of an administrative action option for handling the detected discrepancy to curate the information corpus. ## Claims ### Claim 1 A method comprising the steps of: producing, by a processor, on a user interface device, a parse tree illustration that includes a cluster model and one or more confidence ratings, wherein the cluster model contains one or more previously-detected discrepancies between at least a first portion of a first electronic document and at least a second electronic document in a curated information corpus, wherein the first portion and the second document comprise natural language text strings from a plurality of natural language text documents related to the curated information corpus, and wherein the one or more confidence ratings are previously-assigned to the natural language text strings from the plurality of natural language text documents; responsive to receipt, by a processor from a user via the user interface device, of a user selection of at least one of the one or more previously-detected discrepancies in the cluster model, returning, to the user via the user interface device, a drill down dialog including one or more drill-down details, wherein the drill-down details include at least two or more conflicting natural language text strings, the one or more confidence ratings, a user-selectable option to reference the parse tree illustration, and at least one user-selectable administrative action option for handling the at least two or more conflicting natural language text strings; and responsive to receipt, by a processor from the user via the user interface device, of a user selection of the at least one user-selectable administrative action option, performing, by the processor, the selected administrative action option to further curate the information corpus and recording, by the processor, the selected administrative action option in a tracking list. ### Claim 2 (depends on 1) The method as set forth in claim 1 wherein the cluster model further comprises confidence factors associated with each discrepancy. ### Claim 3 (depends on 2) The method as set forth in claim 2 wherein confidence factors associated with the selected discrepancy are included in the returned drill-down details. ### Claim 4 A computer program product comprising: a computer-readable memory which is not a propagating signal, per se; and program instructions encoded by the computer-readable memory for causing a processor, when executed, to perform steps comprising: producing, on a user interface device, a parse tree illustration that includes a cluster model and one or more confidence ratings, wherein the cluster model contains one or more previously-detected discrepancies between at least a first portion of a first electronic document and at least a second electronic document in a curated information corpus, wherein the first portion and the second document comprise natural language text strings from a plurality of natural language text documents related to the curated information corpus, and wherein the one or more confidence ratings are previously-assigned to the natural language text strings from the plurality of natural language text documents; responsive to receipt, from a user via the user interface device, of a user selection of at least one of the one or more previously-detected discrepancies in the cluster model, returning, to the user via the user interface device, a drill down dialog including one or more drill-down details, wherein the drill-down details include at least two or more conflicting natural language text strings, the one or more confidence ratings, a user-selectable option to reference the parse tree illustration, and at least one user-selectable administrative action option for handling the at least two or more conflicting natural language text strings; and responsive to receipt, from the user via the user interface device, of a user selection of the at least one user-selectable administrative action option, performing, by the processor, the selected administrative action option to further curate the information corpus and recording, by the processor, the selected administrative action option in a tracking list. ### Claim 5 (depends on 4) The computer program product as set forth in claim 4 wherein the cluster model further comprises confidence factors associated with each discrepancy. ### Claim 6 (depends on 5) The computer program product as set forth in claim 5 wherein confidence factors associated with the selected discrepancy are included in the returned drill-down details. ### Claim 7 A system comprising: a computer processor; a computer-readable memory which is not a propagating signal, per se; and program instructions encoded by the computer-readable memory for causing the processor, when executed, to perform steps comprising: producing, on a user interface device, a parse tree illustration that includes a cluster model and one or more confidence ratings, wherein the cluster model contains one or more previously-detected discrepancies between at least a first portion of a first electronic document and at least a second electronic document in a curated information corpus, wherein the first portion and the second document comprise natural language text strings from a plurality of natural language text documents related to the curated information corpus, and wherein the one or more confidence ratings are previously-assigned to the natural language text strings from the plurality of natural language text documents; responsive to receipt, from a user via the user interface device, of a user selection of at least one of the one or more previously-detected discrepancies in the cluster model, returning, to the user via the user interface device, a drill down dialog including one or more drill-down details, wherein the drill-down details include at least two or more conflicting natural language text strings, the one or more confidence ratings, a user-selectable option to reference the parse tree illustration, and at least one user-selectable administrative action option for handling the at least two or more conflicting natural language text strings; and responsive to receipt, from the user via the user interface device, of a user selection of the at least one user-selectable administrative action option, performing, by the processor, the selected administrative action option to further curate the information corpus and recording, by the processor, the selected administrative action option in a tracking list. ### Claim 8 (depends on 7) The system as set forth in claim 7 wherein the cluster model further comprises confidence factors associated with each discrepancy. ### Claim 9 (depends on 8) The system as set forth in claim 8 wherein confidence factors associated with the selected discrepancy are included in the returned drill-down details.