Efficient Answering of Historical What-if Queries

TL;DR

This paper introduces a new type of database query called historical what-if queries, which assess the impact of hypothetical historical changes on current data, using optimized reenactment techniques for efficiency.

Contribution

The paper presents novel methods for efficiently answering historical what-if queries through reenactment and slicing techniques, enabling actionable insights from hypothetical historical data changes.

Findings

Techniques are effective in reducing reenactment scope.

Implementation in Mahif demonstrates practical efficiency.

Methods enable actionable insights from hypothetical historical scenarios.

Abstract

We introduce historical what-if queries, a novel type of what-if analysis that determines the effect of a hypothetical change to the transactional history of a database. For example, "how would revenue be affected if we would have charged an additional $6 for shipping?" Such queries may lead to more actionable insights than traditional what-if queries as their results can be used to inform future actions, e.g., increasing shipping fees. We develop efficient techniques for answering historical what-if queries, i.e., determining how a modified history affects the current database state. Our techniques are based on reenactment, a replay technique for transactional histories. We optimize this process using program and data slicing techniques that determine which updates and what data can be excluded from reenactment without affecting the result. Using an implementation of our techniques in Mahif (a Middleware for Answering Historical what-IF queries) we demonstrate their effectiveness experimentally.