Web intelligence (WI) is a new direction which can provide a new thought for Web-based problem solving. There are two fundamental issues regarding the effectiveness of using the Web data: mismatch and overload. The mismatch means some interesting and useful data has not been found (or missed out), whereas, the overload means some gathered data is not what users expect.

We argue that the following three steps are necessary to overcome the fundamental issues within WI. The first step is to create an adequate environment (e.g. Semantic Web) for users to manipulate their data (e.g., XML). The second step is the discovery of interesting and useful knowledge from the environment. The third step is data reasoning of discovered knowledge in order to answer what users really want.

Currently, the application of data mining techniques to Web data, called Web mining, is used to discover patterns (knowledge) from Web data. It is indubitable that we can obtain numerous discovered patterns using a Web mining model. However, there still exits a gap between Web mining and the effectiveness of using the Web data, since it is very difficult to use the numerous discovered patterns for dealing with the fundamental issues. One reason is that some discovered patterns might include uncertainties when we extract them. Another reason is that user profiles are changeable.

An ontology-based Web mining model tends to overcome the above difficult problems, which uses ontologies to represent the discovered patterns in order to search the right data for what users want. We have presented a theoretical framework to extend the ontology-based Web mining model. It consists of automatic ontology extraction, reasoning on the ontology and capturing evolving patterns.  The innovation is that the theoretical framework deals with pattern evolution. 

Yuefeng Li

12  Feb 2006

References:

[1] N. Zhong, Representation and construction of ontologies for Web intelligence, International Journal of Foundation of Computer Science, 2002, 13(4): 555-570.

[2] Y. Li and N. Zhong, Ontology-based Web mining model, IEEE/WIC International Conference on Web Intelligence, Canada, 2003, 96-103.

[3] Y. Li and N. Zhong, Capturing evolving patterns for ontology-based Web mining, accepted by  IEEE/WIC/ACM International Conference on Web Intelligence, 20-24 September 2004, Beijing, China, 256-263.

[4] Y. Li and N. Zhong, Web mining model and its applications on information gathering, Knowledge-Based Systems, Vol. 17, 2004, pp 207-217.

[5] Y. Li and N. Zhong, A ontology-based Web mining model, IEEE Transactions on Knowledge and Data Engineering, Vol. 18, No. 4, 2006, pp 554-568.