Online Community Answer Ranking via Syntactic Proposition Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ranking methods for on-line community question answering systems fail to effectively consider the overall quality and diversity of answers, often prioritizing relevance over novelty and diversity, and struggle with condensed, unstructured content typical in CQA settings.
Innovation Solution
The proposed solution involves extracting syntactic propositions from CQA answers, filtering them based on relevance, measuring similarity using semantic approaches, and ranking them using a greedy process that accounts for relevance, novelty, and coverage of important aspects, employing techniques like ESA and CQA-ESA to enhance answer diversity and importance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If typical ranking measures focus on determining whether a particular answer should be selected as a best or more appropriate answer by accounting for qualitative aspects of candidate answers separately, then the relevance to corresponding questions is improved, but the overall quality and diversity of the ranked list of answers deteriorates
Solution Approach 1:
The patent merges multiple ranking measures (relevance, novelty, diversity) into a single comprehensive ranking framework. Instead of treating qualitative aspects separately as typical ranking measures do, the invention combines them into an integrated system that evaluates answers based on multiple dimensions simultaneously, thereby improving both relevance measurement and answer diversity in the ranked list
Solution Approach 2:
The ranking system is designed to perform multiple functions: evaluating relevance, assessing novelty, and measuring diversity all within a single ranking process. This multi-functional approach allows the system to maintain high relevance to questions while simultaneously ensuring diverse and comprehensive answer coverage, resolving the contradiction between precision and versatility
2Reliability
If typical ranking measures prioritize relevance over novelty and diversity, then the importance and relevance to corresponding questions is improved, but the flexibility and variety of ranking measures deteriorates
Solution Approach 1:
The ranking system dynamically adjusts the weight and importance of different measures (relevance, novelty, diversity) based on the specific context and question type. Rather than statically prioritizing relevance, the system can adaptively balance these factors, allowing high reliability when relevance is critical while maintaining flexibility to emphasize novelty or diversity when appropriate for the particular query
Solution Approach 2:
The invention changes the parameters of the ranking system by introducing multiple adjustable parameters that control the contribution of different evaluation criteria. This allows the system to maintain reliable relevance assessment while simultaneously providing flexibility to adjust the balance between relevance, novelty, and diversity based on different ranking scenarios and user needs
Data Source
AI summary
Briefly, example methods, apparatuses, and/or articles of manufacture are disclosed that may be implemented, in whole or in part, using one or more computing devices to facilitate and/or support one or more operations and/or techniques for in part, to facilitate and/or support one or more operations and/or techniques for ranking answers for on-line community questions.