Proceedings: GI 2015

Exploiting analysis history to support collaborative data analysis

Ali Sarvghad , Melanie Tory

Proceedings of Graphics Interface 2015: Halifax, Nova Scotia, Canada, 3 - 5 June 2015, 123-130

DOI 10.20380/GI2015.16

  • Bibtex

    @inproceedings{Sarvghad:2015:10.20380/GI2015.16,
    author = {Sarvghad, Ali and Tory, Melanie},
    title = {Exploiting analysis history to support collaborative data analysis},
    booktitle = {Proceedings of Graphics Interface 2015},
    series = {GI 2015},
    year = {2015},
    issn = {0713-5424},
    isbn = {978-1-4822-6003-8},
    location = {Halifax, Nova Scotia, Canada},
    pages = {123--130},
    numpages = {8},
    doi = {10.20380/GI2015.16},
    publisher = {Canadian Human-Computer Communications Society},
    address = {Toronto, Ontario, Canada},
    }

Abstract

Coordination is critical in distributed collaborative analysis of multidimensional data. Collaborating analysts need to understand what each person has done and what avenues of analysis remain uninvestigated in order to effectively coordinate their efforts. Although visualization history has the potential to communicate such information, common history representations typically show sequential lists of past work, making it difficult to understand the analytic coverage of the data dimension space (i.e. which data dimensions have been investigated and in what combinations). This makes it difficult for collaborating analysts to plan their next steps, particularly when the number of dimensions is large and team members are distributed. We introduce the notion of representing past analysis history from a dimension coverage perspective to enable analysts to see which data dimensions have been explored in which combinations. Through two user studies, we investigated whether 1) a dimension oriented view improves understanding of past coverage information, and 2) the addition of dimension coverage information aids coordination. Our findings demonstrate that a representation of dimension coverage reduces the time required to identify and investigate unexplored regions and increases the accuracy of this understanding. In addition, it results in a larger overall coverage of the dimension space, one element of effective team coordination.