Query Expansion Scoring via Dependency-Based Feature Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The lexical gap between query words and document words in information retrieval systems leads to suboptimal search results, as existing query expansion techniques fail to effectively capture contextual dependencies between terms.
Innovation Solution
The proposed solution involves scoring query alteration candidates based on intra-candidate and query-dependent dependencies using a Markov random field framework, which considers multiple features and translation models to compute a candidate score for selecting relevant expansion terms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If statistical machine translation models are used to rank alteration candidates assuming word independence, then the scoring process is simplified and faster, but the accuracy of capturing contextual dependencies between terms deteriorates
Solution Approach 1:
The patent introduces Markov random fields as an intermediary framework that models dependencies between alteration candidate terms and query terms. This mediator captures contextual relationships that simple word-based models miss, improving accuracy without completely sacrificing computational efficiency through optimized feature computation.
Solution Approach 2:
The patent changes the modeling parameters from independent word assumptions to dependency-based probability models. By incorporating features that capture n-gram dependencies and translation probabilities between query and candidate terms, the system adjusts its parameter representation to reflect contextual relationships more accurately.
2Device complexity
If lexical databases and co-occurrence techniques are used for query expansion, then the process is simpler and more straightforward, but the ability to capture complex contextual relationships between terms deteriorates
Solution Approach 1:
The patent combines multiple scoring features into a composite evaluation framework. Instead of relying on a single simple metric, it integrates translation probabilities, n-gram dependencies, and lexical database information into a unified scoring model that weighs multiple factors to determine candidate quality, thereby improving reliability through diversified evaluation.
Solution Approach 2:
The system incorporates feedback mechanisms where the scoring model learns from the relationships between query terms and candidate terms. By using probabilistic models that capture dependency patterns, the system refines its selection process based on observed contextual relationships, improving relevance through iterative optimization.
Data Source
AI summary
An alteration candidate for a query can be scored. The scoring may include computing one or more query-dependent feature scores and/or one or more intra-candidate dependent feature scores. The computation of the query-dependent feature score(s) can be based on dependencies to multiple query terms from each of one or more alteration terms (i.e., for each of the one or more alteration terms, there can be dependencies to multiple query terms that form at least a portion of the basis for the query-dependent feature score(s)). The computation of the intra-candidate dependent feature score(s) can be based on dependencies between different terms in the alteration candidate. A candidate score can be computed using the query dependent feature score(s) and/or the intra-candidate dependent feature score(s). Additionally, the candidate score can be used in determining whether to select the candidate to expand the query. If selected, the candidate can be used to expand the query.