Score-Based Explanations in Data Management and Machine Learning

TL;DR

This paper explores score-based explanation methods in data management and machine learning, emphasizing causal and counterfactual approaches, and advocates integrating domain knowledge into score computations.

Contribution

It introduces score-based explanation techniques for database query answers and classification results, highlighting the importance of incorporating domain and semantic knowledge.

Findings

Score-based explanations can clarify observed outcomes.

Causal and counterfactual methods are effective for explanations.

Integrating domain knowledge improves explanation relevance.

Abstract

We describe some approaches to explanations for observed outcomes in data management and machine learning. They are based on the assignment of numerical scores to predefined and potentially relevant inputs. More specifically, we consider explanations for query answers in databases, and for results from classification models. The described approaches are mostly of a causal and counterfactual nature. We argue for the need to bring domain and semantic knowledge into score computations; and suggest some ways to do this.