A visual mapping approach for trend identification in multi-attribute data

Jesse Bockstedt, Gediminas Adomavicius

Research output: Contribution to conferencePaperpeer-review

Abstract

Organizations and firms are increasingly capturing more data about their customers, suppliers, competitors, and business environment. Most of this data is multi-attribute (multi-dimensional) and temporal in nature. Data mining and business intelligence techniques are typically used to discover patterns in this data; however, mining meaningful temporal relationships is often difficult. We introduce a new temporal data analysis and visualization technique for representing trends in multi-attribute temporal data using a clustering-based approach. We define a new analytical construct called the temporal cluster graph which maps multi-attribute temporal data into a two-dimensional trend graph that clearly identifies trends in dominant data types over time. We also present C-TREND, a system that implements the proposed technique, and demonstrate applications of technique by analyzing the change in technical characteristics of wireless networking technologies over a six year period.

Original languageEnglish (US)
Pages213-219
Number of pages7
StatePublished - Jan 1 2007
Event17th Workshop on Information Technologies and Systems, WITS 2007 - Montreal, QC, Canada
Duration: Dec 8 2007Dec 9 2007

Other

Other17th Workshop on Information Technologies and Systems, WITS 2007
CountryCanada
CityMontreal, QC
Period12/8/0712/9/07

ASJC Scopus subject areas

  • Information Systems

Fingerprint Dive into the research topics of 'A visual mapping approach for trend identification in multi-attribute data'. Together they form a unique fingerprint.

Cite this