# US 2018/0025406 A1 · Determining recommendations based on user intent

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

- Number: US 2018/0025406 A1
- Document ID: US20180025406A1
- Status: Published application
- Kind: A1
- Application number: 15/670,027
- Filed: 2017-08-07
- Published: 2018-01-25
- Earliest claimed: 2016-07-19
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Manish Choudhary; Srinivasan S. Muthuswamy
- Web page: https://patentgrep.com/patent/US20180025406A1

## CPC

- H04L 67/535: {Tracking the activity of the user (network monitoring arrangements H04L43/00; recording of computer activity G06F11/34)}
- G06N 20/00: Machine learning
- G06Q 30/0201: Market modelling; Market analysis; Collecting market data
- G06Q 30/0256: {User search}
- G06Q 30/0631: {Recommending goods or services}
- H04L 67/02: based on web technology, e.g. hypertext transfer protocol [HTTP]
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- 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

## Abstract

According to one or more embodiments, a method, a computer program product, and a computer system for determining recommendations based on user intent are provided. The method may include identifying, by a server computer, one or more nodes. Weight values may be calculated by the server computer for each of the identified nodes, based on analyzing classes of metadata associated with the identified nodes. A web-browsing history of a user corresponding to the identified nodes may be obtained by the server computer. Based on the obtained web-browsing history, a classification may be determined for the user by the server computer, whereby the classification corresponds to one class of metadata associated with the identified nodes. The server computer may select one or more of the identified nodes having a weight value greater than a predetermined threshold value, whereby the selected nodes correspond to the determined classification of the user.

## Claims

### Claim 1

A computer-implemented method for determining recommendations based on user intent, the method comprising:
identifying one or more nodes, wherein the nodes represent products available for purchase;
identifying edges between each of the identified nodes, wherein the identified edges have an initial weight value of zero;
determining nodes selected by a user from among the identified nodes;
determining nodes viewed by the user associated with each of the nodes selected by the user, wherein the nodes viewed by the user includes only those nodes viewed by the user prior to the user's selection of nodes from among the identified nodes;
incrementing the weight value for all edges between nodes selected by the user and nodes viewed by the user;
calculating a weight value for each node selected by the user, wherein the calculated weight is a sum total of the weight values of all edges associated with each node selected by the user;
obtaining a web-browsing history of the user corresponding to the identified nodes;
organizing (grouping) the identified nodes into groups (classes) based on characteristics identified by analyzing metadata associated with the identified nodes;
identifying one group of the organized groups corresponding to the user based on the obtained web-browsing history;
selecting one or more of the identified nodes from the identified group having a weight value greater than a predetermined threshold value; and
displaying the one or more selected nodes to the user.