Search Result Ranking via Similar Query Data Borrowing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines face challenges in ranking search results effectively, as they often lack sufficient user behavior data for queries, leading to suboptimal relevance in search results, especially when queries are similar but not identical.
Innovation Solution
The method involves scoring and selecting similar historical queries based on their similarity to a user-submitted query, using weights for terms such as exact matches, variants, and stopwords, and combining user behavior data from these queries to refine the ranking of search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user behavior data is used for ranking search results, then relevance of search results is improved, but insufficient data for specific queries leads to suboptimal ranking
Solution Approach 1:
The patent combines user behavior data from multiple similar historical queries to create a composite quality of result statistic for the current query. By merging data from queries with similar terms and structures, the system overcomes the limitation of insufficient data for any single query while maintaining ranking relevance.
Solution Approach 2:
The patent introduces similar historical queries as intermediaries between the current query and the user behavior data repository. These intermediate queries serve as bridges to access relevant user behavior data when direct data for the current query is insufficient, enabling indirect data retrieval that improves ranking accuracy.
2Measurement precision
If search engines rely on exact query matching for ranking, then data accuracy is improved, but similarity between queries is not充分利用, leading to lost ranking opportunities
Solution Approach 1:
The patent changes the matching parameter from exact query equality to similarity-based matching. By computing quality of result statistics based on similar queries rather than requiring exact matches, the system maintains data accuracy while becoming adaptable to query variations, thereby utilizing query similarity to improve ranking.
Solution Approach 2:
The patent makes the ranking system universal by enabling it to handle both exact matches and similar queries through a unified similarity-based approach. The system can process diverse query forms (different phrasings, synonyms, related terms) using the same mechanism, improving versatility without sacrificing accuracy.
Data Source
AI summary
In general, in one aspect, embodiments of the invention feature a computer-implemented method for providing input to a document ranking process for ranking a plurality of documents, the document ranking process taking as input a quality of result statistic for an individual document in the plurality of documents, the plurality of documents having been identified in response to a query, the quality of results statistic being for the query and the individual document. The method for providing input can include, for a first document identified as a search result of a user-submitted query, scoring one or more other queries based upon a similarity to the user-submitted query, each of the one or more other queries being different from the user-submitted query.