Why not?

  • Authors:

  • Adriane Chapman;H. V. Jagadish

  • Affiliations:

  • The MITRE Corporation, McLean, VA, USA;University of Michigan, Ann Arbor, MI, USA

  • Venue:
  • Proceedings of the 2009 ACM SIGMOD International Conference on Management of data
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

As humans, we have expectations for the results of any action, e.g. we expect at least one student to be returned when we query a university database for student records. When these expectations are not met, traditional database users often explore datasets via a series of slightly altered SQL queries. Yet most database access is via limited interfaces that deprive end users of the ability to alter their query in any way to garner better understanding of the dataset and result set. Users are unable to question why a particular data item is Not in the result set of a given query. In this work, we develop a model for answers to WHY NOT? queries. We show through a user study the usefulness of our answers, and describe two algorithms for finding the manipulation that discarded the data item of interest. Moreover, we work through two different methods for tracing the discarded data item that can be used with either algorithm. Using our algorithms, it is feasible for users to find the manipulation that excluded the data item of interest, and can eliminate the need for exhausting debugging.