Data Management Technology for Interactive Vis


In the last two decades, interactive visualization and analysis have become a central tool in data-driven decision making. Concurrently to the contributions in data visualization, the data management community has spent significant effort in developing technology that directly benefits interactive analysis. We contribute a systematic review of this adjacent field, and highlight a number of techniques and principles we believe to be underappreciated in visualization work. We seek to bring attention to: materialized views, approximate query processing, user modeling and query prediction, multi-query optimization, provenance techniques, and indexing techniques.

By a combination of use cases and taxonomies from the visualization literature, we show that these ideas can positively change how we visualization designers build our systems, while also highlighting research gaps, where future work is necessary.


A Structured Review of Data Management Technology for Interactive Visualization and Analysis. Leilani Battle and Carlos Scheidegger. IEEE TVCG (VIS 2020), to appear. pdf

Interactive Website

If you want to get a feel for what’s in the bibliography, or even contribute new articles to it, please visit the interactive version of the bibliography.

Methodology and Materials

We conducted a systematic literature review of six different venues going back to at least 2000 (three data management venues: ICDE, SIGMOD and VLDB and three visualization venues: IEEE VIS, ACM CHI, and EuroVis).

We categorized the results we obtained by coding the publications we found along two axes. The first axis characterizes the interaction techniques which are applicable for the techniques proposed; we obtained the codes themselves from combining two well-regarded task taxonomies from the literature. The second axis characterizes the kind of technical work, which we obtained by a combination of the keywords used by the publications and the our experience as authors of working in the intersection of the data management and visualization fields.

The result of this methodology is a bibliography annotated with interaction codesencode, select, navigate, arrange, change, filter, aggregate, annotate, derive and record — as well as implementation codesmaterialized-view, aqp (for “approximate query processing”), mqo (for “multi-query optimization”), user-modeling, and index (for “indexing-based techniques”). In addition, as we have worked thrhough the identified articles, we have collected all other surveys under a separate category (coded as survey).

The resulting bibliography has 104 articles and is available as a BibTeX file.


Prof. Battle is partially funded by NSF grant IIS-1850115, CRII: CHS: Modeling Analysis Behavior to Support Interactive Exploration of Massive Datasets.

Prof. Scheidegger is partially funded by NSF grant IIS-1815238, III: Small: An end-to-end pipeline for interactive visual analysis of big data.


This material is based upon work supported or partially supported by the National Science Foundation under Grant Number 1815238, project titled “III: Small: An end-to-end pipeline for interactive visual analysis of big data”, and under Grant Number 1850115, project titled “CRII: CHS: Modeling Analysis Behavior to Support Interactive Exploration of Massive Datasets”.

Any opinions, findings, and conclusions or recommendations expressed in this project are those of author(s) and do not necessarily reflect the views of the National Science Foundation.

Web page last update: 2020-09-11.