Analytic Trails: Supporting Provenance, Collaboration, and Reuse for Visual Data Analysis by Business Users

Abstract

In this paper, we discuss the use of analytic trails to support the needs of business users when conducting visual data analysis, focusing particularly on the aspects of analytic provenance, asynchronous collaboration, and reuse of analyses. We present a prototype implementation of analytic trail technology as part of Smarter Decisions ( a web-based visual analytic tool, with the goal of helping business users derive insights from structured and unstructured data. To understand the value and shortcomings of trails in supporting visual analytic tasks in business environments, we performed a user study with 21 participants. While the majority of participants found trails to be useful for capturing and understanding the provenance of an analysis, they viewed trails as more valuable for personal use rather than for communicating the analytic process to other people as part of a collaboration. Study results also indicate that rich search mechanisms for easily finding relevant trails (or portions of a trail) is critical to the successful adaptation and reuse of existing saved trails.