Random Forests Model of Data Mining Classification Used to Identify Unknown Authorship of Tamil Articles

Abstract

Salford Predictive Modeler (SPM) developed many data mining tools. In this research paper, Random Forest (RF) model classification has recently been used as new tool which involves extensive research in various disciplines. Application of RF method has increased considerably in areas of data mining and classification problems in the field of Text Mining over the most recent decade. In this research paper, an attempt is completed to deal with the authorship attribution problems using a RF model for qualifying the articles of unknown authorship to one of the existing writers of the same period. A set of stylistic variables such as function words is made use of for classification purposes. After that the results of authorship attribution are discussed. The database is based on blocks of articles, in which each block contains many sentences. Initially, the three scholars writing styles are identified in the previous studies, and they had three different writing styles. The group of articles of anonymous authorship to the articles written by contemporary Tamil scholars of the similar period, namely Mahakavi Bharathiar (MB), Subramaniya Iyer (SI) and T. V. Kalyanasundaram (TVK) is dealt in the current research paper. The above said three accepted scholars had written a number of articles on India's Freedom Movement during the pre-independence time. Originally, all the three writers contributed their articles by attributing their names. The oppressive attitude of the then British administration forced all the three patriots to write articles on the same theme for unknown publications without mentioning their names.