Classifying and aggregating context attributes for business service requests - No 'One-Size-Fits-All'

Rebecca Copeland, Noel Crespi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

When building decision-making models from disparate observations, there are no set rules to guide the designer on how to organize available information, how to classify vital aspects, how to emphasize important ones in the aggregation process, and how to deal with conflict and uncertainty in the aggregation procedures. This paper draws on the experience of structuring a business and risk model that evaluates service requests, which requires not only dynamic context-based decisions, but also situational and behavioral perspectives, with high uncertainty and wide variations of attribute styles. This study focuses on design issues that affect classification and aggregation options, such as corroboration, primacy and discord, and provides examples of classified key-factors that demonstrate the design issues. The paper suggests procedures and algorithms to fit the design, but shows that there is no universal method - there is no 'one-size-fits-all'.

Original languageEnglish
Title of host publicationProceedings - 2014 IEEE International Congress on Big Data, BigData Congress 2014
EditorsPeter Chen, Peter Chen, Hemant Jain
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages808-815
Number of pages8
ISBN (Electronic)9781479950577
DOIs
Publication statusPublished - 22 Sept 2014
Externally publishedYes
Event3rd IEEE International Congress on Big Data, BigData Congress 2014 - Anchorage, United States
Duration: 27 Jun 20142 Jul 2014

Publication series

NameProceedings - 2014 IEEE International Congress on Big Data, BigData Congress 2014

Conference

Conference3rd IEEE International Congress on Big Data, BigData Congress 2014
Country/TerritoryUnited States
CityAnchorage
Period27/06/142/07/14

Keywords

  • Aggregation
  • CART
  • Classification
  • Corroboration
  • Credibility
  • DST
  • Discordant
  • MCDM
  • OWA
  • SAW
  • WPM
  • policy

Fingerprint

Dive into the research topics of 'Classifying and aggregating context attributes for business service requests - No 'One-Size-Fits-All''. Together they form a unique fingerprint.

Cite this