Clustering of cancer text documents in the medical field using machine learning heuristics

Abstract

The data clustering over medical text documents plays a major role in extracting relevant information from the documents. However, most of the methods fails in finding the accurate solution on finding the relevant cancer type due to the presence of redundant data items. It is hence necessary to develop a clustering framework that strictly eliminates the redundant data items. In this paper, we present a clustering framework that tends to accurately cluster the cancer text documents to predict what type of cancer is present in a patient. A large database is tested and clustering using the machine learning model. The clustering framework consists of pre-processing the text documents, feature extraction, feature selection and clustering. The clustering using multi-support vector machine enables optimal clustering of text documents. The cancer datasets is used to validate the models over various medline cancer documents dataset. The experimental validation shows improved clustering of documents using the proposed models than other methods.