FEDEX: An Explainability Framework for Data Exploration Steps

TL;DR

FEDEX is a framework that helps data scientists understand why certain data points are interesting during exploration by analyzing their contribution to overall data insights, using semantic relatedness to improve explanations.

Contribution

It introduces a novel method for identifying and explaining interesting data subsets based on their contribution to column-wise interestingness, incorporating semantic relatedness for better explanations.

Findings

Effective identification of interesting data subsets

Improved explanations using semantic relatedness

Validated on multiple real-world datasets

Abstract

When exploring a new dataset, Data Scientists often apply analysis queries, look for insights in the resulting dataframe, and repeat to apply further queries. We propose in this paper a novel solution that assists data scientists in this laborious process. In a nutshell, our solution pinpoints the most interesting (sets of) rows in each obtained dataframe. Uniquely, our definition of interest is based on the contribution of each row to the interestingness of different columns of the entire dataframe, which, in turn, is defined using standard measures such as diversity and exceptionality. Intuitively, interesting rows are ones that explain why (some column of) the analysis query result is interesting as a whole. Rows are correlated in their contribution and so the interesting score for a set of rows may not be directly computed based on that of individual rows. We address the resulting computational challenge by restricting attention to semantically-related sets, based on multiple notions of semantic relatedness; these sets serve as more informative explanations. Our experimental study across multiple real-world datasets shows the usefulness of our system in various scenarios.