SHARQ: Explainability Framework for Association Rules on Relational Data

2024-12-24

AI Summary

Existing association rules in relational data lack interpretability and fail to quantify individual or collective contributions of data elements to rule validity. Method: This paper introduces SHARQ, the first Shapley-value-based framework for quantifying association rule importance. It innovatively integrates Shapley value theory with combinatorial optimization and association rule mining to enable efficient, exact computation of both single-element and multi-element importance, extending evaluation to both rule-level and attribute-level granularity. We design a quasi-linear-time algorithm for single-element importance and a low-amortized-cost algorithm for multi-element importance. Contribution/Results: Evaluated on a novel benchmark comprising 45 real-world rule sets, SHARQ significantly outperforms baseline methods, demonstrating strong theoretical rigor and practical computational efficiency.

Technology Category

• Data Mining & Knowledge Management: Rule Mining & Pattern Mining
• Knowledge Representation and Reasoning: Qualitative Reasoning
• Machine Learning: Learning Preferences or Rankings

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Association rules are an important technique for gaining insights over large relational datasets consisting of tuples of elements (i.e. attribute-value pairs). However, it is difficult to explain the relative importance of data elements with respect to the rules in which they appear. This paper develops a measure of an element's contribution to a set of association rules based on Shapley values, denoted SHARQ (ShApley Rules Quantification). As is the case with many Shapely-based computations, the cost of a naive calculation of the score is exponential in the number of elements. To that end, we present an efficient framework for computing the exact SharQ value of a single element whose running time is practically linear in the number of rules. Going one step further, we develop an efficient multi-element SHARQ algorithm which amortizes the cost of the single element SHARQ calculation over a set of elements. Based on the definition of SHARQ for elements we describe two additional use cases for association rules explainability: rule importance and attribute importance. Extensive experiments over a novel benchmark dataset containing 45 instances of mined rule sets show the effectiveness of our approach.

Problem

• Importance Quantification
• Association Rules
• Big Data Analysis

Innovation

• SHARQ system
• Shapley value
• efficiency in large-scale data