User Attribute Relevance Scoring for Content Recommendations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computer systems that generate online content recommendations suffer from data ambiguity, leading to relevant content being downgraded, resulting in excessive resource consumption and user inefficiency due to the display of irrelevant content.
Innovation Solution
A computer system computes relevance scores for user attributes based on recent online activity, using a sliding window of time and inverse frequency discounting to accurately reflect current user interests, thereby improving the relevance of content recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If computer systems generate recommendations based on user data without disambiguation, then content can be displayed quickly, but relevant content is downgraded in favor of irrelevant content
Solution Approach 1:
The patent segments user data into multiple attributes (e.g., job title, skills, education, experience) and calculates separate relevance scores for each attribute. This segmentation allows the system to evaluate different dimensions of user interest independently, improving the precision of relevance assessment while maintaining efficient processing through modular computation.
Solution Approach 2:
The system dynamically changes the weighting parameters of different user attributes based on their relevance scores. By adjusting these parameters in real-time according to user activity and profile information, the system can accurately prioritize relevant content while maintaining efficient recommendation generation through parameter-based control.
2Loss of time
If computer systems display irrelevant content to users, then more content can be shown, but users spend longer time searching and navigating
Solution Approach 1:
The system performs preliminary disambiguation and relevance assessment before content is displayed to users. By pre-calculating relevance scores for different content items based on user attributes and activity, the system filters and prioritizes content in advance, eliminating the need for users to spend time searching through irrelevant content while maintaining an appropriate volume of recommended items.
Solution Approach 2:
The system uses user activity data (clicks, views, interactions) as feedback to continuously refine relevance scores and content prioritization. This feedback mechanism allows the system to learn from user behavior patterns and improve content ranking over time, reducing user search time while maintaining optimal content volume through adaptive refinement.
3Loss of energy
If computer systems generate and display irrelevant content, then recommendation coverage is increased, but processing power and computational resources are wasted
Solution Approach 1:
The system extracts and focuses computational resources on calculating relevance scores for only the most important user attributes and content items. By identifying and processing only the critical subset of data that contributes most to recommendation quality, the system reduces processing power consumption while maintaining comprehensive recommendation coverage through selective extraction of key information.
Solution Approach 2:
The system dynamically adjusts computational parameters such as relevance score thresholds and attribute weighting to optimize the balance between processing efficiency and recommendation coverage. By changing these parameters based on system load and user profile complexity, the system can reduce energy consumption during peak loads while maintaining versatile recommendation coverage when resources are available.
4Measurement precision
If computer systems use static attribute analysis for recommendations, then system complexity is reduced, but accuracy of relevance prediction is insufficient
Solution Approach 1:
The patent divides the recommendation system into separate modular components, each responsible for calculating relevance scores for specific user attributes (e.g., one module for job title matching, another for skills, another for education). This segmentation improves prediction accuracy by allowing specialized processing for each attribute type while keeping individual module complexity manageable through functional decomposition.
Solution Approach 2:
The system implements a universal relevance scoring framework that can handle multiple attribute types and content categories through a common computational approach. This multi-functional architecture improves prediction accuracy across diverse recommendation scenarios while reducing overall system complexity by reusing the same core algorithms and data structures across different recommendation contexts.
Data Source
AI summary
Techniques for using online user activity in determining relevance of attributes to improve computer functionality in generating recommendations of online content are disclosed herein. In some embodiments, a computer system calculates a corresponding relevance score for each attribute of a user based on a total number of online postings for which the user has performed at least one of a plurality of online actions within a particular sliding window of time defining a most recent time period, an attribute activity number representing a number of online postings in the plurality of online postings that have the attribute, and an inverse of a frequency value representing how many of a total number of online postings published within the particular sliding window of time have the attribute. In some embodiments, the computer system causes at least one recommendation associated with the user to be displayed based on the calculated relevance scores.