Context-Weighted Content Ranking Using Implicit Influence Relationships
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing content ranking systems often bias search results towards older or more reviewed content, neglecting newer or more relevant content due to lack of direct relationships, leading to suboptimal user experiences.
Innovation Solution
Derive implicit influence relationships from user download sequences, campaign data, and content review data to create a graph with context-dependent weights, adjusting the importance of influence and similarity relationships based on user intent and context.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If search results are ranked based on traditional link structure and review data, then older and more reviewed content is prioritized, but newer and potentially more relevant content is neglected
Solution Approach 1:
The patent applies dynamics by making the graph weights context-dependent rather than static. The weights of influence relationships are dynamically adjusted based on contextual signals from the search query, allowing the ranking system to adapt to different search intents and prioritize newer content when appropriate while maintaining reliability for established content when needed
Solution Approach 2:
The patent changes the parameter of graph edge weights from fixed values to context-dependent values. By modifying the weight parameters based on contextual analysis of search queries and content characteristics, the system can balance between prioritizing established content and promoting newer relevant content, resolving the contradiction between reliability and timeliness
2Reliability
If implicit influence relationships are derived from multiple data sources, then content relevance is improved, but system complexity increases
Solution Approach 1:
The patent uses an intermediary graph structure with context-dependent weights to mediate between multiple data sources and the final ranking output. This graph serves as an intermediary layer that integrates information from user download sequences, campaign data, and content review data while managing complexity through a unified weighted relationship model
Solution Approach 2:
The patent creates a universal graph-based ranking system that can process multiple types of data sources (user download sequences, campaign data, review data) through a single unified framework. This multi-functional approach improves content relevance accuracy while avoiding the need for separate processing systems for each data type
Data Source
AI summary
The present technology has the ability to establish connections between content that do not have direct or explicit relationships. Implicit influence relationships can be established from user download sequence data, campaign data with keyword targeting, and content review data that mentions other content. Using these influence relationships, the relevance of content items can be determined based on the influence relationship of linked content items and a similarity relationship of content items. However, the importance of the influence relationship in ranking content items can vary depending on the parameters against which the content item is considered relevant. To address this, the present technology includes a context-driven factor that is used as a weight to adjust the impact of the influence relationship of the ranking, depending on the parameters against which the content item is considered relevant.