KDD Methodology/Algorithms

In recent years, worldwide interest in Knowledge Discovery and Data Mining (KDD) has soared as it has wide applicability and the emergence of bigger data collection in many contexts provides us with many challenges from efficiency of analysis to privacy concerns.  We have worked on a number of algorithms to fit within different stages of the KDD process.

One current concern is Privacy Preserving Data Mining (PPDM). We have worked on PDM by developing data perturbation methods that enable the data owners to transform the original data at the pre-processing stage so that it can be shared.  The perturbation method should guarantee the privacy of the data while preserving much of the data utility for analysis.  You can see how we have adapted non-metric Multi-Dimensional Scaling (MDS) for PPDM.

Additionally, we have applied expertise in optimization, particularly Multi-Objective Optimization(MOO) to create efficient data mining algorithms for both classification and clustering.  Our more recent work on clustering with MOO can be explored.

We are currently also exploring clustering of complex objects, i.e. objects that may have associated structure, unstructured and semi structured data as well as multimedia data. 

Past Projects

We have made a number of contributions in techniques for data mining and KDD in the last 10 years including:

References

  1. Hills, J, Bagnall, A, De La Iglesia, B and Richards, G (2013) BruteSuppression: a size reduction method for Apriori rule sets. Journal of Intelligent Information Systems. ISSN 0925-9902
  2. Reynolds, A and de la Iglesia, B (2009) A multi-objective GRASP for partial classification. Soft Computing - A Fusion of Foundations, Methodologies and Applications, 13 (3). pp. 227-243.
  3. Alotaibi, Khaled and De La Iglesia, Beatriz (2013) Privacy-Preserving SVM Classification using Non-metric MDS. In: SECURWARE 2013, the Seventh International Conference on Emerging Security Information, Systems and Technologies. IARIA, pp. 30-35.
  4. Alotaibi, K., Rayward-Smith, V.J., Wang, W. and De La Iglesia, B. (2012) Non-linear dimensionality reduction for privacy-preserving data classification. In: Proceedings - 2012 ASE/IEEE International Conference on Privacy, Security, Risk and Trust and 2012 ASE/IEEE International Conference on Social Computing, SocialCom/PASSAT 2012. UNSPECIFIED, pp. 694-701. ISBN 9780769548487
  5. Alotaibi, K, Rayward-Smith, V and De La Iglesia, B (2011) Non-metric Multidimensional Scaling for Privacy-Preserving Data Clustering. In: Intelligent Data Engineering and Automated Learning - IDEAL 2011. Spring, pp. 287-298.
  6. Kirkland, O, Rayward-Smith, V and De La Iglesia, B (2011) A Novel Multi-Objective Genetic Algorithm for Clustering. In: Intelligent Data Engineering and Automated Learning - IDEAL 2011. Springer, pp. 317-326.

Research Team

Dr. Beatriz de la Iglesia, Oliver Kirkland, Khaled Alotaibi, Aalaa Mojahed