Data Mining in E-Learning

This chapter presents an innovative approach for performing data mining on documents, which serves as a basis for knowledge extraction in e-learning environments. The approach is based on a radical model of text data that considers phrasal features paramount in documents, and employs graph theory to facilitate phrase representation and efficient matching. In the process of text mining, a grouping (clustering) approach is also employed to identify groups of documents such that each group represents a different topic in the underlying document collection. Document groups are tagged with topic labels through unsupervised key-phrase extraction from the document clusters. The approach serves in solving some of the difficult problems in e-learning where the volume of data could be overwhelming for the learner, such as automatically organizing documents and articles based on topics, and providing summaries for documents and groups of documents.1