Jincy B. Chrystal and Stephy Joseph, IIITM-K, India

Text mining and Text classification are the two prominent and challenging tasks in the field of Machine learning. Text mining refers to the process of deriving high quality and relevant information from text, while Text classification deals with the categorization of text documents into different classes. The real challenge in these areas is to address the problems like handling large text corpora, similarity of words in text documents, and association of text documents with a subset of class categories. The feature extraction and classification of such text documents require an efficient machine learning algorithm which performs automatic text classification. This paper describes the classification of product review documents as a multi-label classification scenario and addresses the problem using Structured Support Vector Machine. The work also explains the flexibility and performance of the proposed approach for efficient text classification.

Text classification, Multi-label classification, Structured Support Vector Machine