Content Relevancy Model for Stale Data Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Companies face difficulties in identifying and managing stale or outdated digital content, which can impact search results and user satisfaction, due to the large volume of content and limited time for review.

Innovation Solution

A system and method for content management that collects and analyzes document and user data to build a content relevancy model, using topic modeling and heatmaps to determine action states for content, allowing for dynamic adjustments such as updating or retiring content to maintain freshness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual review of digital content is performed to identify stale content, then content relevancy can be maintained, but the process becomes inefficient and time-consuming due to large content volume

Engineering Contradiction:

Improvecontent relevancyVSAvoidcontent review efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent replaces manual mechanical review processes with automated machine learning models and natural language processing systems. The system automatically analyzes document data, access patterns, and user interactions to identify stale content, eliminating the need for human reviewers to manually assess each piece of content while maintaining high accuracy in detecting relevancy issues.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables content to be automatically assessed and flagged for staleness through self-service mechanisms. The machine learning models continuously monitor content characteristics and access patterns, allowing the content management system to self-identify stale content without external human intervention, thereby improving review efficiency while maintaining reliability.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If comprehensive content analysis is performed to ensure accuracy in identifying stale content, then measurement precision improves, but the computational complexity and time required increase

Engineering Contradiction:

Improvestale content detection accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the content analysis process into multiple independent components: document data analysis, access data processing, user interaction tracking, and machine learning model evaluation. Each component handles a specific aspect of content assessment, allowing the system to achieve comprehensive analysis through modular processing rather than monolithic complex analysis, thereby improving precision while managing complexity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary analysis of content characteristics and access patterns before conducting full staleness assessment. Machine learning models are pre-trained on historical data and continuously refined, allowing the system to quickly evaluate new content against established patterns. This preliminary preparation enables accurate detection without requiring complex real-time computation for each assessment.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If frequent content updates are performed to maintain freshness, then content relevancy improves, but the loss of time for content management increases

Engineering Contradiction:

Improvecontent freshnessVSAvoidcontent management time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements periodic monitoring and assessment of content staleness using machine learning models that continuously analyze access patterns and document characteristics. Instead of manual frequent updates, the system performs automated periodic evaluations at optimal intervals, triggering updates only when staleness thresholds are exceeded. This periodic automated action maintains content freshness while minimizing the time investment required compared to manual frequent review and update cycles.

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS11734586B2Detecting and improving content relevancy in large content management systems

Publication Date: 2023.08.22 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method, a computer system, and a computer program product for managing content relevancy is provided. Embodiments of the present invention may include collecting and analyzing a plurality of data, wherein the plurality of data includes document data, document access data and user data. Embodiments of the present invention may include retrieving topic model content based on the plurality of data. Embodiments of the present invention may include building a machine learning (ML) model to determine one or more topics contained in the topic model content. Embodiments of the present invention may include generating a heatmap based on the user data. Embodiments of the present invention may include building a content relevancy model (CRM) based on the ML model and the heatmap. Embodiments of the present invention may include determining an action state for the document data. Embodiments of the present invention may include storing the CRM.