Neural Network Answer Quality Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating the quality of answers on internet platforms face challenges in accurately distinguishing high-quality answers from low-quality ones due to limitations in linguistic analysis tools and reliance on user voting, which can be influenced by time and group psychology, leading to inefficiencies and inaccuracies.
Innovation Solution
A method that extracts question and answer feature expressions using neural networks to determine textual quality and semantic correlation, enabling a quality score to be calculated based on both aspects, thereby accurately distinguishing high-quality answers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If user voting is used to evaluate answer quality, then the evaluation process is simple to implement, but the accuracy is reduced due to influence from time and group psychology factors
Solution Approach 1:
The patent introduces a neural network-based quality evaluation model as an intermediary between user voting and final answer ranking. This model processes both the voting data and textual features to produce a corrected quality score, eliminating direct reliance on flawed user voting while maintaining implementation simplicity through automated computation
Solution Approach 2:
The patent replaces the mechanical user voting system with a neural network-based automated evaluation system. The neural network processes textual features and voting patterns to generate quality scores, substituting human psychological influences with algorithmic computation that is objective and consistent
2Ease of manufacture
If traditional linguistic analysis tools are used to evaluate answer quality, then the implementation is straightforward, but the ability to accurately distinguish high-quality from low-quality answers is limited
Solution Approach 1:
The patent changes the evaluation parameters from traditional linguistic features to neural network-derived semantic representations. By transforming the input space to include contextual embeddings and semantic correlations, the system achieves superior quality distinction while maintaining implementation ease through standardized neural network architectures
3Device complexity
If only textual quality is considered in answer evaluation, then the evaluation process is simple, but the ability to assess semantic relevance to the question is insufficient
Solution Approach 1:
The patent segments the evaluation process into two independent modules: textual quality evaluation and semantic correlation evaluation. Each module focuses on specific aspects (language quality vs. question-answer relevance), reducing overall complexity while achieving comprehensive and precise assessment through combination of both modules
Data Source
AI summary
Embodiments of the present disclosure provide a method, an apparatus, a device and a storage medium for evaluating quality of an answer. The method includes extracting a question feature expression of a question and an answer feature expression of an answer with respect to the question, the question and the answer being represented in a form of text; determining a measurement of textual quality of the answer based on the answer feature expression; determining a measurement of correlation on semantics between the question and the answer based on the question feature expression and the answer feature expression; and determining a quality score of the answer with respect the question based on the measurement of textual quality and the measurement of correlations. Therefore, a high-quality answer may be accurately obtained.