Content Recommendation System Using Multi-Technique Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional approaches to recommending content items in social networks often fail to provide users with content that is of high interest, especially as the size of social networks grows, leading to a challenge in identifying and presenting targeted content related to a user's expressed interests.
Innovation Solution
The system employs a combination of techniques such as comments-based, token-based, and tag-based methods to identify follow-up content items related to an original content item, applying constraints like time differences and weight values, and utilizes machine learning to optimize the selection and presentation of follow-up content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional content recommendation approaches are used in growing social networks, then the volume of available content increases, but the ability to identify content items of high interest to the user deteriorates
Solution Approach 1:
The patent segments the content identification process into multiple distinct techniques: comments-based technique (analyzing user comments and links), token-based technique (tokenizing content and comparing representations using tf-idf and cosine similarity), and tag-based technique (using hierarchical tags and categories). Each technique processes content differently to identify follow-up items, allowing the system to handle large volumes while maintaining identification precision through diversified analysis approaches
Solution Approach 2:
The patent applies multiple parameters and constraints to filter and rank content items, including time difference thresholds (filtering content published within a specific time window), weight value thresholds (assigning weights to different content attributes), and similarity thresholds (using cosine similarity scores). These parameter changes enable precise identification of relevant content amidst large volumes by dynamically adjusting filtering criteria
2Reliability
If multiple techniques and constraints are applied to identify follow-up content items, then the relevance of recommended content improves, but the system complexity increases
Solution Approach 1:
The system divides the complex content recommendation task into three separate, modular techniques (comments-based, token-based, and tag-based), each handling a specific aspect of content analysis. This segmentation allows each module to be developed, optimized, and maintained independently while contributing to the overall recommendation reliability through their combined output
Solution Approach 2:
The patent creates a universal content analysis framework that can process and analyze content through multiple techniques simultaneously. The system applies the same constraint mechanisms (time thresholds, weight thresholds, similarity thresholds) across all three techniques, providing a multi-functional approach that handles diverse content types and user interactions through a unified system architecture
Data Source
AI summary
Systems, methods, and non-transitory computer readable media configured to detect access by a user to an original content item relating to a story. At least one of a comments based technique, a token based technique, and a tag based technique is performed on content items. Constraints are applied to identify at least one follow up content item from the content items relating to a development of the story.