Matching Search Queries to TV Subtitles for Real-Time Interest Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing second screen applications struggle to provide timely and relevant information to users while watching television, as they rely on inaccurate audience predictions or costly live polls, and fail to quickly identify user interest in specific TV content.
Innovation Solution
The implementation matches search queries from users to television subtitles in real-time, identifying spikes in query frequency to correlate with specific events in TV programs, and provides relevant search results to users through a second screen application.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If audience prediction methods are used to identify user interest, then the system can provide information without live polling, but the precision and relevance of the information decreases
Solution Approach 1:
The patent introduces search queries as an intermediary signal to infer user interest. Instead of directly polling users or relying on demographic predictions, the system uses search queries submitted by users as a mediator to understand what topics or events in the TV program are capturing their attention. This intermediary approach provides real-time, actionable data about user interest without requiring direct user input or invasive polling.
Solution Approach 2:
The system implements feedback by continuously monitoring search queries and using them to adjust and improve information delivery. Search queries serve as feedback signals that indicate what users are interested in, allowing the system to refine its understanding of user preferences and provide more relevant information over time. This creates a closed-loop system where user behavior directly informs future information delivery.
2Loss of information
If live polls are conducted to gather user feedback, then relevant information can be obtained, but the cost and time required increases significantly
Solution Approach 1:
The system leverages search queries that users voluntarily submit as self-service feedback. Users are not burdened with completing polls or providing direct feedback; instead, their natural search behavior automatically generates the feedback data the system needs. This eliminates the need for time-consuming poll distribution and processing while still gathering comprehensive information about user interests and preferences.
Solution Approach 2:
The system uses search queries as a proxy or copy of direct user feedback. Rather than requiring users to explicitly state their interests through polls, the system analyzes search queries as indirect copies of user intent. This copying approach captures user interest information efficiently without the overhead of direct feedback collection mechanisms.
3Ease of operation
If demographic-based audience prediction is used, then information can be provided without real-time data, but the relevance to individual users decreases
Solution Approach 1:
The system transitions from static demographic-based predictions to dynamic, real-time adaptation using search query data. By continuously monitoring search queries and adjusting information delivery based on current user interest signals, the system becomes adaptable to changing user preferences and contexts. This dynamic approach allows the system to personalize information delivery in real-time while maintaining operational simplicity through automated query analysis.
Data Source
AI summary
A process identifies a search query spike from queries submitted by users during a first span of time, which is less than a predefined duration. The spike corresponds to a set of queries identified as equivalent. The frequency of submitting queries from the set during the first time span exceeds the frequency of submitting queries from the set during an average span of time. The process correlates the spike to a broadcast video program by matching terms from the set of search queries to subtitle terms appearing in the video program at a first location. The first location in the video program was broadcast within a predefined time before the first span of time. The process receives notification from a user device indicating user interest in the video program. The process transmits to the user device search results corresponding to some search queries from the set of search queries.