Trending Term Detection via Multi-Source Engagement Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for detecting trending topics rely solely on search logs, which may not accurately capture users' interests and focus, as other forms of user engagement with content are ignored.
Innovation Solution
A method and system that identify trending terms by gathering information from diverse sources, including search logs, content items, and engagement metrics, to compute a trendiness score using a scoring model, thereby selecting trending terms based on comprehensive user engagement analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional search log-based methods are used to detect trending topics, then the detection process is simple, but the accuracy of trending topic detection deteriorates because other forms of user engagement are ignored
Solution Approach 1:
The patent combines multiple data sources including search logs, content items, and engagement metrics into a unified analysis framework. This merging of previously separate data streams enables comprehensive trending topic detection that captures both explicit search behavior and implicit engagement patterns, thereby improving detection accuracy while managing system complexity through integrated processing.
Solution Approach 2:
The detection system is designed to handle multiple types of data (search queries, content metadata, engagement metrics) through a unified scoring model that evaluates trendiness across diverse input types. This multi-functional approach allows the same system to process different data categories and generate comprehensive trending topic rankings, improving accuracy without requiring separate specialized systems for each data type.
2Loss of information
If only search logs are used for trending detection, then data processing is straightforward, but user engagement analysis becomes incomplete
Solution Approach 1:
The patent merges search log data with engagement metrics and content information into a unified data structure that preserves all user interaction types. This combination ensures no user engagement information is lost, as the system captures both explicit search behavior and implicit engagement patterns such as time spent on content and interaction depth, while processing them through an integrated pipeline.
Solution Approach 2:
The patent segments user engagement into multiple measurable dimensions including search frequency, content viewing time, interaction depth, and engagement recency. This segmentation allows the system to process complex engagement data in manageable components, each contributing to the overall trendiness score, thereby reducing processing complexity while maintaining information completeness.
3Measurement precision
If diverse data sources are integrated for trending term identification, then trendiness scoring accuracy improves, but information processing time increases
Solution Approach 1:
The patent implements a time-window based approach that processes only recent engagement data and search logs within a defined period, rather than analyzing all historical data. This partial action strategy maintains trendiness scoring accuracy by focusing on current user behavior patterns while significantly reducing processing time by excluding outdated information that no longer reflects current trends.
Solution Approach 2:
The patent segments the data processing into distinct stages: data collection from multiple sources, feature extraction from segmented data groups, trendiness scoring based on extracted features, and ranking generation. This segmentation allows parallel processing of different data types and enables optimization at each stage, reducing overall processing time while maintaining comprehensive analysis accuracy.
Data Source
AI summary
The present teaching relates to trending term identification. Information in different categories from different sources associated with each of terms being evaluated for trendiness is obtained within a recent period and linked to generate a data group for each term. Features are extracted for each term based on information in the corresponding data group and are used to compute a trendiness score in accordance with a scoring model. Trending terms are selected from the terms based on their trendiness scores.