Expectation Maximization Algorithm Model for Authorship Attribution in Tamil Article Database

• Published in: Bulletin of pure & applied sciences. Sec. E, Mathematics & statistics Vol. 35e; no. 1; pp. 19 - 30
• Main Authors: Priya, R. Lakshmi, Manimannan, G
• Format: Journal Article
• Language: English
• Published: BPAS Research 01.01.2016
• Subjects: Authorship Attribution; Classification; Clustering; EM algorithm and Function Words; Stylometry; Tamil Articles
• ISSN: 0970-6577; 2320-3226
• DOI: 10.5958/2320-3226.2016.00003.5

This research article attempts to classify the distinguish authorship problem of three existing Tamil scholars during the Indian independence period, namely, Mahakavi Bharathiar (MB), Subramaniya Iyer (SI), and T. V. Kalyanasundaranar (TVK) using Expectation Maximization (EM) algorithm method. The three trendy scholars stated above had written more number of articles on India's Freedom Movement during the pre-independence period and published in the magazine called, India in the year 1906 is considered for this study. The database consists of thirty six articles of blocks of MB, thirty one blocks of articles of TVK, thirty eight blocks of articles of SI, each block having many sentences and nineteen function words parameters are used. In the above point of view, this study attempts to identify three authorship of three authors using their stylistic traits and classification of articles of MB, TVK and SI using EM algorithm model. EM algorithm has been used newly as a modern tool and attracted broad research in various disciplines including Text mining. Application of EM algorithm has increased considerably in areas of pattern recognition and classification problems in the field of Text mining over the last decade.