# US 2025/0147983 A1 · Scalable topological summary construction using landmark point selection

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0147983 A1
- Document ID: US20250147983A1
- Status: Published application
- Kind: A1
- Application number: 18/671,263
- Filed: 2024-05-22
- Published: 2025-05-08
- Earliest claimed: 2014-03-05
- Assignees: SymphonyAI Sensa LLC
- Applicants: SymphonyAI Sensa LLC
- Inventors: Gurjeet Singh; Lawrence Spracklen; Ryan Hsu
- Web page: https://patentgrep.com/patent/US20250147983A1

## CPC

- G06F 16/285: {Clustering or classification}
- G06F 16/29: Geographical information databases
- G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)}
- G06Q 10/08355: {Routing methods}
- G16B 5/00: ICT specially adapted for modelling or simulations in systems biology, e.g. gene-regulatory networks, protein interaction networks or metabolic networks
- G16B 40/00: ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY

## Abstract

An example method comprises receiving data points, determining at least one size of a plurality of subsets based on a constraint of at least one computation device or an analysis server, transferring each of the subsets to different computation devices, each computation device selecting a group of data points to generate a first sub-subset of landmarks, add non-landmark data points that have the farthest distance to the closest landmark to create an expanded sub-subset of landmarks, create an analysis landmark set based on a combination of expanded sub-subsets of expanded landmarks from different computation devices, perform a similarity function on the analysis landmark set, generate a cover of the mathematical reference space to create overlapping subsets, cluster the mapped landmark points based on the overlapping subsets, create a plurality of nodes, each node being based on the clustering, each landmark point being a member of at least one node.

## Claims

### Claim 1

A method comprising:
receiving a large number of data points;
determining at least one size of a plurality of subsets of the large number of data points based on constraints of at least one of a plurality of computation devices or an analysis server, each data point of the large number of data points being a member of at least one of the plurality of subsets of the large number of data points;
transferring each of the plurality of subsets of large number of data points to a respective one of the plurality of computation devices;
for each of the plurality of subsets of data points by an associated computation device of the plurality of computation devices:
selecting, by the associated computation device, a group of data points from the subset of data points to generate a first sub-subset of landmarks;
adding, by the associated computation device, a non-landmark data point of the subset of data points to the first sub-subset of landmarks to create an expanded sub-subset of landmarks, adding the non-landmark data points comprising:
calculating first data point distances between each non-landmark data point and each landmark;
identifying a shortest data point distance from among the first data point distances for each non-landmark data point;
identifying a particular non-landmark data point with a longest first landmark distance of all the shortest data path distances; and
adding the particular non-landmark data point to the first sub-subset of landmarks to expand the first sub-subset of landmarks to generate an expanded set of landmarks until the expanded sub-subset of the expanded landmarks reaches a predetermined number of members, repeat adding the non-landmark data points;
creating an analysis landmark set based on a combination of expanded sub-subsets of expanded landmarks;
performing a similarity function on the analysis landmark set to map landmark points of the analysis landmark set to a mathematical reference space;
generating a cover of the mathematical reference space to divide the mathematical reference space into overlapping subsets;
clustering the mapped landmark points of the analysis landmark set based on the overlapping subsets of the cover in the mathematical reference space;
creating a plurality of nodes, each of the plurality of nodes being based on the clustering of the mapped landmark points of the analysis landmark set, each landmark point of the analysis landmark set being a member of at least one node; and
connecting at least two of the plurality of nodes with an edge if the at least two of the plurality of nodes share at least one landmark point of the analysis landmark set as a member.
