Mining Relational Databases

For relational database, data is made up of tuples in relations. A query language on a relational database describes a selection of tuples from relations and hence a pattern. Queries are well suited for describing patterns and the frequency of a query can be easily measured by the number of results the query returns after evaluation against the relational database. Conjunctive queries, that are described in a sub-language of the standard relational query language SQL, satisfy the demands of good mining patterns and hence form a good basis for applying data mining on arbitrary relational databases. Although conjunctive queries have been studied extensively in the context of relational databases they form a difficult class from the perspective of data mining. Our goal is to study interesting subclasses of conjunctive queries and the patterns they can represent, and develop efficient algorithms to mine them. Using this we can apply data mining to relational databases in an efficient and intuitive way.

In studying the different interesting subclasses of conjunctive queries we wel explore there potential theoretically and experimentally. Experimentally we will develop efficient algorithms. To start we will focus on frequent pattern mining and association rule mining (finding correlations between patterns). To guarantee the efficiency we will take into account the important factors of data mining algorithms like the size of the search space, duplicates, redundancy, monotonicity, constraints and many more important factors that require a redefinition in the context of conjunctive queries.

2009

• Attended the 15th ACM SIGKDD Conference On Knowledge Discovery and Data Mining (KDD-09), Paris, France, June 28th to Juli 1st, 2009

2008

• Organised The 19th European Conference on Machine Learning (ECML 2008) and the 12th European Conference on Principles and Practice of Knowledge Discovery in Databases (PKDD 2008), Antwerp, Belgium September 15th to 19th, 2008
• Presented a poster at the 2008 WOG Seminar on Logic and Computation, Oostduinkerke, Belgium, April 29th to 30th, 2008
• Presented at the SIAM International Conference on Data Mining (SDM08), Atlanta, Gorgia, USA,  April 24th to 26th, 2008
• Attended the Dutch-Belgian Database Day, Namur, Belgium, October 10th, 2008
  

2007

• Attended The 18th European Conference on Machine Learning (ECML 2007) and the 11th European Conference on Principles and Practice of Knowledge Discovery in Databases (PKDD 2007), Warsaw, Poland, September 17th to 21st, 2007
• Presented at the Dutch-Belgian Database Day, Eindhoven, Nederland, November 29th, 2007
  

2006

• Attended the Joint APrIL/IQ Workshop, Titisee, Germany, March 15th to 18th, 2006

The courses I have taught in the past.

• Databases
• Project Databases
  
• Machines en Berekenbaarheid (together with Jeroen Avonts)