Live Text Broadcasting Using Historical Sentiment Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine-based live text broadcasting methods for sporting events provide dull and factual statements, leading to poor broadcasting quality and low viewer engagement, as they lack human-like sentiment and cadence.
Innovation Solution
A method and device that generate real-time utterances for live broadcasting by using historical live broadcasting utterances with human-like sentiment, incorporating topic marking and sentiment grading to create engaging and emotive descriptions of match data, simulating a human-like cadence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine-based live text broadcasting is used, then broadcasting efficiency is improved and human resources are saved, but the broadcasting quality becomes plain and dull with poor viewer engagement
Solution Approach 1:
The system copies the stylistic characteristics, sentiment expressions, and cadence patterns from historical manual broadcasting utterances to generate automated broadcasting content. By replicating the linguistic features and emotional tone of human broadcasters through text processing and template matching, the system achieves both high efficiency and human-like quality in live text broadcasting
Solution Approach 2:
The system transforms plain match data into enriched broadcasting utterances by adjusting parameters such as sentiment intensity, cadence rhythm, and descriptive depth. Historical utterances provide reference parameter ranges for sentiment grading and cadence control, allowing the system to dynamically adjust output characteristics to match human broadcasting standards while maintaining automated efficiency
2Reliability
If manual typing-based live text broadcasting is used, then broadcasting quality with human-like sentiment is achieved, but human resources are consumed and efficiency is reduced
Solution Approach 1:
The system enables automated broadcasting to serve itself by using historical manual broadcasting utterances as training data and reference templates. The automated system processes match data through sentiment analysis and cadence modeling to generate human-like utterances independently, eliminating the need for continuous manual intervention while maintaining broadcasting quality
Solution Approach 2:
The system performs preliminary analysis of historical broadcasting utterances to extract sentiment patterns, cadence characteristics, and expression templates before live broadcasting begins. This pre-processing creates a repository of stylistic reference data that enables the automated system to generate high-quality utterances in real-time without manual typing during the actual broadcast
Data Source
AI summary
The disclosed embodiments provide a method and device for live text broadcasting of a match. The method for live text broadcasting of a match comprises: receiving match data for a current match in real time; generating real-time utterances for live broadcasting used for describing the current match data with more human-like sentiment according to historical live broadcasting utterances used for describing match data in historical matches with human-like sentiment; and outputting the real-time live broadcasting utterances. The disclosed embodiments can conduct live text broadcasting of a match with human-like cadence, thus improving the quality of the live text broadcasting of a match.