Provectories: Embedding-Based Analysis of Interaction Provenance Data

Abstract

Understanding user behavior patterns and visual analysis strategies is a long-standing challenge. Existing approaches rely largely on time-consuming manual processes such as interviews and the analysis of observational data. While it is technically possible to capture a history of user interactions and application states, it remains difficult to extract and describe analysis strategies based on interaction provenance. In this paper, we propose a novel visual approach to the meta-analysis of interaction provenance. We capture single and multiple user sessions as graphs of high-dimensional application states. Our meta-analysis is based on two different types of two-dimensional embeddings of these high-dimensional states: layouts based on (i) topology and (ii) attribute similarity. We applied these visualization approaches to synthetic and real user provenance data captured in two user studies. From our visualizations, we were able to extract patterns for data types and analytical reasoning strategies.

The Influence of Visual Provenance Representations on Strategies in a Collaborative Hand-off Data Analysis Scenario

“…While modern techniques still implement examples from both data coverage and interaction timelines [60,66,75], prior work has seen greater emphasis on using machine learning to extract patterns and assist in the summarization of time [24,27,59,74] and less emphasis on comparative studies investigating the implications of different provenance summaries with people. Questions remain about how best to summarize interaction histories in digestible ways that help analysts by balancing content and cognitive load.…”

Preliminary Perspectives on Information Passing in the Intelligence Community

“…For example, recommender systems can help analysts find relevant information they might not know to look for [4], and natural language text summarization can greatly reduce the time needed for human interpretation of large collections of text [5]. Complementary advancements in capabilities for identifying patterns, anomalies, and relationships among entities offer the promise of revealing intelligence findings that might otherwise have been missed [6][7][8]. Given these developments, the grand challenge of the Summer Conference on Applied Data Science (SCADS), an 8-week program hosted by the Laboratory for Analytic Sciences (LAS) at North Carolina State University, is to use AI/ML to generate tailored daily reports (TLDR) for intelligence analysts to increase the efficiency and efficacy of their work [9].…”