Enhancing Algebraic Query Relaxation with Semantic Similarity

TL;DR

This paper introduces an algebraic framework for query relaxation in cooperative databases, incorporating semantic similarity heuristics to improve relevance assessment of answers.

Contribution

It presents an algebraic formulation of query relaxation operators and proposes heuristics for assigning similarity degrees to answer tuples.

Findings

Algebraic operators for query relaxation are formalized.

Heuristics for semantic similarity scoring are developed.

Enhanced relevance determination for query answers.

Abstract

Cooperative database systems support a database user by searching for answers that are closely related to his query and hence are informative answers. Common operators to relax the user query are Dropping Condition, Anti-Instantiation and Goal Replacement. In this article, we provide an algebraic version of these operators. Moreover we propose some heuristics to assign a degree of similarity to each tuple of an answer table; this degree can help the user to determine whether this answer is relevant for him or not.