A Taxonomy of Dirty Time-Oriented Data

Teaser Image
Author
Conference Paper
Editor
Abstract
Data quality is a vital topic for business analytics in order to gain accurate insight and make correct decisions in many data-intensive industries. Albeit systematic approaches to categorize, detect, and avoid data quality problems exist, the special characteristics of time-oriented data are hardly considered. However, time is an important data dimension with distinct characteristics which a ffords special consideration in the context of dirty data. Building upon existing taxonomies of general data quality problems, we address `dirty' time-oriented data, i.e., time-oriented data with potential quality problems. In particular, we investigated empirically derived problems that emerge with di fferent types of time-oriented data (e.g., time points, time intervals) and provide various examples of quality problems of time-oriented data. By providing categorized information related to existing taxonomies, we establish a basis for further research in the field of dirty time-oriented data, and for the formulation of essential quality checks when preprocessing time-oriented data.
Keywords
Year of Publication
2012
Conference Name
Lecture Notes in Computer Science (LNCS 7465): Multidisciplinary Research and Practice for Information Systems (Proceedings of the CD-ARES 2012)
Publisher
Springer, Berlin / Heidelberg
Conference Location
Prague, Czech Republic
ISBN Number
978-3-642-32497-0
DOI
reposiTUm Handle
Funding projects
Paper
Download citation