# US 2026/0093901 A1 · Text clustering with heuristic and multi-metric control > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2026/0093901 A1 - Document ID: US20260093901A1 - Status: Published application - Kind: A1 - Application number: 19/414,619 - Filed: 2025-12-10 - Published: 2026-04-02 - Earliest claimed: 2023-07-26 - Assignees: JPMorgan Chase Bank, N.A. - Applicants: JPMorgan Chase Bank, N.A. - Inventors: Jialei SHI; John F. BUFORD - Web page: https://patentgrep.com/patent/US20260093901A1 ## CPC - G06F 40/166: Editing, e.g. inserting or deleting - G06F 40/151: Transformation - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Implementations generally relate to text clustering with heuristic and multi-metric control. In some implementations, a method includes receiving an electronic source document containing text. The method further includes dividing the text into text units, encoding the text units, and transforming the text units into numerical values. The method further includes generating a graph of the text units based on the numeric values, where the graph includes nodes corresponding to the text units and edges corresponding to pairs of the text units. The method further includes ordering the text units into text clusters based on the graph of the text units. The method further includes generating an electronic target document that presents the text clusters based on one or more preference heuristics. ## Claims ### Claim 1 A system comprising: one or more processors; and logic encoded in one or more non-transitory computer-readable storage media for execution by the one or more processors and when executed operable to cause the one or more processors to perform operations comprising: receiving an electronic source document containing text; dividing the text into text units; encoding the text units; transforming the text units into numerical values, wherein the text units are associated with their respective numeric values; and generating a graph of the text units based on the numeric values, wherein the graph comprises nodes corresponding to the text units and edges corresponding to pairs of the text units. ### Claim 2 (depends on 1) The system of claim 1, the logic when executed is further operable to cause the one or more processors to perform operations comprising: ordering the text units into text clusters based on the graph of the text units; and generating an electronic target document that presents the text clusters based on one or more preference heuristics. ### Claim 3 (depends on 1) The system of claim 1, the logic when executed is further operable to cause the one or more processors to perform operations comprising selecting one or more preference heuristics, wherein the one or more preference heuristics comprise balancing. ### Claim 4 (depends on 1) The system of claim 1, the logic when executed is further operable to cause the one or more processors to perform operations comprising selecting one or more one or more preference heuristics, wherein the one or more preference heuristics comprise coherence. ### Claim 5 (depends on 1) The system of claim 1, wherein when generating the graph, the logic when executed is further operable to cause the one or more processors to perform operations comprising associating one or more preference heuristics with the nodes. ### Claim 6 (depends on 1) The system of claim 1, wherein when generating the graph, the logic when executed is further operable to cause the one or more processors to perform operations comprising computing distance values for the edges. ### Claim 7 (depends on 1) The system of claim 1, wherein when generating the graph, the logic when executed is further operable to cause the one or more processors to perform operations comprising computing seed clusters from the text units based on one or more preference heuristics associated with the nodes. ### Claim 8 A non-transitory computer-readable storage medium with program instructions stored thereon, the program instructions when executed by one or more processors are operable to cause the one or more processors to perform operations comprising: receiving an electronic source document containing text; dividing the text into text units; encoding the text units; transforming the text units into numerical values that are associated with the text units; and generating a graph of the text units based on the numeric values, wherein the graph comprises nodes corresponding to the text units and edges corresponding to pairs of the text units. ### Claim 9 (depends on 8) The computer-readable storage medium of claim 8, the instructions when executed are further operable to cause the one or more processors to perform operations comprising: ordering the text units into text clusters based on the graph of the text units; and generating an electronic target document that presents the text clusters based on one or more preference heuristics. ### Claim 10 (depends on 8) The computer-readable storage medium of claim 8, the instructions when executed are further operable to cause the one or more processors to perform operations comprising selecting one or more preference heuristics, wherein the one or more preference heuristics comprise balancing. ### Claim 11 (depends on 8) The computer-readable storage medium of claim 8, the instructions when executed are further operable to cause the one or more processors to perform operations comprising selecting one or more one or more preference heuristics, wherein the one or more preference heuristics comprise coherence. ### Claim 12 (depends on 8) The computer-readable storage medium of claim 8, wherein when generating the graph, the instructions when executed are further operable to cause the one or more processors to perform operations comprising associating one or more preference heuristics with the nodes. ### Claim 13 (depends on 8) The computer-readable storage medium of claim 8, wherein when generating the graph, the instructions when executed are further operable to cause the one or more processors to perform operations comprising computing distance values for the edges. ### Claim 14 (depends on 8) The computer-readable storage medium of claim 8, wherein when generating the graph, the instructions when executed are further operable to cause the one or more processors to perform operations comprising computing seed clusters from the text units based on one or more preference heuristics associated with the nodes. ### Claim 15 A computer-implemented method comprising: receiving an electronic source document containing text; dividing the text into text units; encoding the text units; transforming the text units into numerical values; and generating a graph of the text units based on the numeric values, wherein the graph comprises nodes corresponding to the text units and edges corresponding to pairs of the text units. ### Claim 16 (depends on 15) The method of claim 15, further comprising: ordering the text units into text clusters based on the graph of the text units; and generating an electronic target document that presents the text clusters based on one or more preference heuristics. ### Claim 17 (depends on 15) The method of claim 15, further comprising selecting one or more preference heuristics, wherein the one or more preference heuristics comprise balancing. ### Claim 18 (depends on 15) The method of claim 15, further comprising selecting one or more preference heuristics, wherein the one or more preference heuristics comprise coherence. ### Claim 19 (depends on 15) The method of claim 15, wherein when generating the graph, the method further comprises associating one or more preference heuristics with the nodes. ### Claim 20 (depends on 15) The method of claim 15, wherein when generating the graph, the method further comprises computing distance values for the edges.