Search Engine Short Answer Scoring via Passage Consensus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines that support short answers often produce varying quality answers due to the lack of consideration for the accuracy of the selected passages, leading to misleading or incorrect answers.
Innovation Solution
An improved scoring engine, or accuracy score prediction engine, is used to determine the quality of short answers by scoring candidate passages based on consensus with other passages from different search results, ensuring that only accurate answers are displayed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If short answers are displayed based on top-ranked search results without accuracy scoring, then the search engine can quickly provide answers to factual queries, but the quality and accuracy of the answers deteriorates
Solution Approach 1:
The accuracy score prediction engine performs preliminary scoring of candidate passages from top-ranked search results before they are displayed as short answers. This preliminary action filters out inaccurate passages while maintaining the speed advantage of using top-ranked results, resolving the contradiction between quick answer provision and answer accuracy.
Solution Approach 2:
The system uses feedback from accuracy score prediction to adjust which passages are displayed as short answers. By continuously evaluating and scoring passages against the query, the system ensures that only accurate passages are displayed while maintaining the rapid response capability of the search engine.
2Reliability
If multiple passages are scored to determine accuracy, then the accuracy of short answers improves, but the processing time and computational complexity increases
Solution Approach 1:
The system extracts and scores only the most relevant candidate passages from top-ranked search results rather than processing all passages. By selectively extracting and evaluating only the most promising candidates, the system reduces computational complexity while maintaining high accuracy in short answer selection.
Solution Approach 2:
The accuracy score prediction engine acts as an intermediary between the search results and the short answer display. This intermediary component simplifies the overall process by providing a standardized scoring mechanism that evaluates passages against the query, reducing the complexity of determining accurate answers without requiring complex multi-pass analysis.
Data Source
AI summary
Techniques of generating short answers for queries by a search engine include performing a training operation on a corpus of training data to train the score prediction engine, the corpus of training data including candidate passages providing short answers for display in callouts and remaining respective passages, from which a top scoring short answer is generated. In such implementations, the corpus of training data further includes the remaining respective passages and the respective titles of the candidate passage and remaining respective passages.