Latent User Community Detection via Transient Topic Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Topology-based community detection methods fail to identify communities of users with similar conceptual interests due to users having similar interests but not being explicitly connected, and social connections being influenced by factors other than interest similarity.

Innovation Solution

A method that determines communities of users with similar temporal behavior by extracting transient topics from user-generated electronic content, modeling these topics as collections of highly correlated semantic concepts, and using graph partitioning to identify user communities based on their contributions to these topics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If topology-based community detection methods are used, then explicit user connections are utilized for community identification, but users with similar interests but no explicit connections cannot be identified

Engineering Contradiction:

Improvecommunity identification accuracyVSAvoidability to identify latent communities

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent introduces electronic content as an intermediary element that mediates between users and community detection. Instead of directly analyzing user connections, the system analyzes content generated by users, which serves as a proxy for user interests and behaviors. This intermediary approach enables identification of users with similar interests even when no explicit connections exist between them.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical system of topology-based detection (which relies on explicit user connections and graph structures) with a content-based analysis system. By substituting the connection-based mechanism with content-based mechanism, the system can detect communities through semantic similarity in electronic content rather than through structural connections in the social network.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If topology-based community detection methods are used, then social connection structure is analyzed, but connections influenced by non-interest factors such as friendship and kinship cannot be distinguished

Engineering Contradiction:

Improveinterest similarity detectionVSAvoidanalysis method complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts and analyzes only the content-related aspects of user behavior by examining electronic content generated by users. This extraction approach separates interest-based signals from non-interest-based social connections, allowing the system to focus specifically on identifying users with similar conceptual interests without being confounded by friendship, kinship, or other non-interest factors.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the analysis parameter from social connection structure to electronic content characteristics. By shifting the parameter being analyzed from topological relationships to content-based features (such as topics, themes, and semantic elements), the system can detect interest similarity more accurately while avoiding the complexity of distinguishing between different types of social connections.

Inventive Principle:

Principle #35Parameter changes

3Reliability

If traditional community detection methods are used, then static user relationships are analyzed, but temporal dynamics of user interests cannot be captured

Engineering Contradiction:

Improvetemporal behavior identificationVSAvoidtime period analysis duration

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent applies periodic action by dividing the time period into discrete intervals and analyzing electronic content generated during each interval. This periodic analysis approach enables the system to capture temporal dynamics of user interests by examining how content characteristics evolve over time, identifying transient topics and temporal patterns in user behavior that static methods would miss.

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS10885131B2System and method for temporal identification of latent user communities using electronic content

Publication Date: 2021.01.05 BAGHERI EBRAHIM

AI summary

Various embodiments are described herein for a system and method for determining a community of users with similar temporal behaviour from a plurality of users that generate electronic content during a time period by, for example, accessing the electronic content from a data store using a processing unit; determining at least one transient topic from the accessed electronic content for the time period using a topic extractor; determining contributions of the users to the identified at least one transient topic using a user community detector; determining the community of users as the users that have similar temporal contributions to the at least one identified transient topic using the user community detector; and providing a recommendation based on a determined user community.