Related Program Extraction for Accurate TV Rating Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to predict rating points for target programs with high accuracy, necessitating improved methods for extracting related programs for accurate prediction.
Innovation Solution
An information processing apparatus that acquires past programs as candidates for rating point prediction of a target program, and extracts related programs based on broadcast station, date, time, and content using techniques like editing distance and BERT-based vectorization to enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to extract related programs for rating point prediction, then the process is simple, but the prediction accuracy is insufficient
Solution Approach 1:
The program extraction process is divided into multiple independent modules: acquisition unit for collecting candidate programs, extraction unit for identifying related programs using multiple criteria (broadcast station, date, time, program name, content), and prediction unit for calculating rating points. This segmentation allows each module to specialize in specific tasks, improving overall prediction accuracy while maintaining manageable system complexity through modular design.
Solution Approach 2:
The system employs multiple parameters for program matching including broadcast station, broadcast date and time, program name, and program content. By changing and combining multiple parameters simultaneously, the system achieves more precise identification of related programs, thereby improving rating point prediction accuracy without requiring overly complex algorithms.
2Measurement precision
If multiple criteria are used to extract related programs, then prediction accuracy improves, but processing complexity increases
Solution Approach 1:
The extraction process is segmented into distinct evaluation dimensions: broadcast station matching, temporal matching (date and time), program name comparison, and content analysis. Each dimension is processed independently by dedicated functional units, allowing the system to handle multiple criteria simultaneously while keeping the complexity of each individual criterion manageable through specialization.
3Reliability
If comprehensive program data is analyzed, then prediction reliability increases, but processing time increases
Solution Approach 1:
The acquisition unit pre-collects and stores candidate program data including broadcast station, date, time, program name, and content information before the prediction process begins. This preliminary action ensures that comprehensive data is readily available when needed, improving prediction reliability without requiring time-consuming data collection during the actual prediction execution.
Solution Approach 2:
The analysis process is divided into parallel segments that can process different aspects of program data simultaneously: metadata comparison (station, date, time, name) and content analysis. This segmentation enables comprehensive data analysis to be performed more efficiently, reducing overall processing time while maintaining high prediction reliability through thorough evaluation of all program characteristics.
Data Source
AI summary
Provided are an information processing apparatus, an information processing method, and an information processing program capable of extracting a related program to be used for rating point prediction of a target program with high accuracy. An information processing apparatus (30) according to one aspect of the present disclosure includes an acquisition unit (31), an extraction unit (32), and a prediction unit (33). The acquisition unit (31) acquires past programs that are candidates for a related program to be used for rating point prediction of the target program. The extraction unit (32) extracts the related program from among the past programs based on a broadcast station, a broadcast date and time, a program name, and a program content of the past program. The prediction unit (33) predicts a rating point of the target program by using the related program.