Off to a Good Start: Using Clustering to Select the Initial Training Set in Active Learning

Rong Hu, Dublin Institute of Technology
Brian Mac Namee, Dublin Institute of Technology
Sarah Jane Delany, Dublin Institute of Technology

Document Type Conference Paper

Proceedings of the Twenty-Third International Florida Artificial Intelligence Research Society Conference (FLAIRS 2010)


Active learning (AL) is used in textual classification to alleviate the cost of labelling documents for training. An important issue in AL is the selection of a representative sample of documents to label for the initial training set that seeds the process, and clustering techniques have been successfully used in this regard. However, the clustering techniques used are nondeterministic which causes inconsistent behaviour in the AL process. In this paper we first illustrate the problems associated with using non-deterministic clustering for initial training set selection in AL. We then examine the performance of three deterministic clustering techniques for this task and show that performance comparable to the non-deterministic approaches can be achieved without variations in behaviour.