SMART software for decision makers KDD experience

https://doi.org/10.1016/S0950-7051(01)00170-8Get rights and content

Abstract

SMART software for decision makers (SSDM) was a Department of Trade and Industry (DTI) initiative running during the period 1998–2000. The Centre for Adaptive Systems at the University of Sunderland was appointed by the DTI to run one of the two demonstrator clubs in the UK. The purpose of these clubs was to facilitate technology transfer between academia and industry, in the areas of fault diagnosis and prediction, intelligent control systems, and knowledge discovery in databases (KDD).

Sunderland SSDM Club decided to use the various industrial members' problems and data to build a number of ‘mini-demonstrators’. In the KDD cluster, three demonstrator applications were developed, accompanied by supporting material and a series of seminars, which illustrated the various stages in the KDD process to all club members.

This paper describes three KDD application demonstrators, developed with data from a manufacturing company, with consultants in business clustering, and from data from a local police force, to investigate the phenomena of repeat victimization.

The work involved data pre-processing, data transformation, data mining, and the development of visual tools for interpretation. With both the business clustering and police data much of the time was spent in data preparation, and so tools were developed so that the members could conduct their own data mining and interpretation experiments, lessening the need for the extraction of domain knowledge from the members.

Introduction

The Centre for Adaptive Systems at the University of Sunderland is a focussed research group affording industry the opportunity to achieve real benefits from advanced computing techniques in areas that include (among others) condition monitoring, intelligent control, and knowledge discovery in databases (KDD). The Centre was appointed by the Department of Trade and Industry (DTI) to run one of the SMART software for decision makers (SSDM) demonstrator clubs precisely to transfer this technological expertise to industry.

Three separate demonstrator applications were developed in the following areas: (i) detection of duplications in a parts database—this work was solely directed at data pre-processing; (ii) investigation of business clusters—this work included data pre-processing, transformation and interpretation; and (iii) investigation of repeat victimization in a crimes database—this work included data pre-processing, transformation, data mining and interpretation.

Because of the time scale involved for the SSDM project, approximately 3 months was dedicated to each of the demonstrator systems. It is known that determining the business objective or question is the key to the data mining process [1]. While each company had an objective, obtaining the relevant domain knowledge to support this was not possible due to the time scale, and so SSDM focussed on the development of visualization tools (see, for instance [2], [3], [4]) that the various companies could take away and further experiment with. This is in accord with the observation that data mining can and should be packaged in such a way that the business professionals can participate directly in the data mining process [5].

This paper describes KDD technology transfer between the Centre for Adaptive Systems and three very different companies. The manufacturing company required help with data cleaning, the business clustering consultants required help with data cleaning, data transformation and data visualization, and the police force were interested in the phenomena of repeat victimization. In two of the projects visualization tools were developed so that the business professionals could participate directly in the data mining process [5].

Section snippets

Conclusions

By means of data and problems supplied by SSDM Club members, and presentation of results in seminars, SSDM was able to illustrate the KDD process and the use of a wide range of algorithms. All the data presented to SSDM required a lot of pre-processing and preparation. The time scale involved meant that while simple experiments could be carried out, the inclusion (or extraction) of sufficient domain knowledge meant that software tools were developed for the individual companies to conduct their

Acknowledgements

The University of Sunderland is grateful to the DTI for the grant ‘SMART software for decision makers’ which made this work possible.

References (34)

  • U.M. Fayyad et al.

    Data mining and KDD: promises and challenges

    Future Gen. Comput. Syst.

    (1997)
  • T.F. Smith et al.

    J. Mol. Biol.

    (1981)
  • S.B. Needleman et al.

    J. Mol. Biol.

    (1970)
  • M.W. Craven et al.

    Using neural networks for data mining

    Future Gen. Comput. Syst.

    (1997)
  • Visible Decisions (Online), Available: http://www.pinetreecapital.com/pine0-2-13.html, January 17,...
  • SpotFire (Online), Available: http://www.spotfire.com/, January 17,...
  • Metaphor Mixer (Online), Available: http://www.inworldvr.com/partners/maxus/mixer.html, January 17,...
  • A. Montgomery, Data mining: computer support discovering and deploying best practice in business and public service....
  • S. Wermter, G. Arevian, C. Panchev, Recurrent neural network learning for text-routing. In: Proceedings of the...
  • S. Wermter, G. Arevian, C. Panchev, Hybrid neural plausibility networks for news agents. In: Proceedings of the...
  • P. Baldi et al.

    Bioinformatics: Machine Learning Approach

    (1998)
  • D. Gusfield

    Algorithms on Strings, Trees and Sequences

    (1997)
  • A. Brazma et al.

    Approaches to the automatic discovery of patterns in biosequences

    J. Comput. Biol.

    (1998)
  • R. Durbin et al.

    Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids

    (1998)
  • W.R. Pearson

    CABIOS

    (1997)
  • M. Waterman

    Introduction to Computational Biology

    (1995)
  • W.R. Pearson et al.

    Proc. Natl Acad. Sci. USA

    (1988)
  • Cited by (6)

    • An intelligent decision-support model using FSOM and rule extraction for crime prevention

      2010, Expert Systems with Applications
      Citation Excerpt :

      The results confirm that the association method can benefit to both the efficiency and accuracy of mining. Other related works include (Corcoran, Wilson, & Ware, 2003; Oatley & Ewart, 2003; Oatley, MacIntyre, Ewart, & Mugambi, 2002; Wang et al., 2002; Xue & Brown, 2003). These all indicate that the discovery of each crime trend with association knowledge provides a promising solution to the crime problem.

    • Crimes analysis software: 'Pins in maps', clustering and Bayes net prediction

      2003, Expert Systems with Applications
      Citation Excerpt :

      When building decision support systems in any application domain the vital questions are ‘who is the intended user, and what is his/her aim.’ This becomes very pertinent in criminal investigations, as the role of the prospective user will influences the design of the system (Oatley, MacIntyre, Ewart, & Mugambi, 2002), for instance, Police and other intelligence agents will want to find the culprits (based upon ‘reasonable suspicion’), while the prosecutors will want to convict the suspect (‘beyond reasonable doubt’) (Zeleznikow, 2002). The most commonly available software for decision support of a Police officer would be a Geographical Information System (GIS).

    • Computer applications for handling legal evidence, police investigation and case argumentation

      2012, Computer Applications for Handling Legal Evidence, Police Investigation and Case Argumentation
    View full text