Comparison and analysis of the machine learning in the movie subtitle document classification

Abstract

With the rapid iteration of film and technology, the quantity of movie related subtitle document has been rapidly expanded. However, the issue of classifying these documents is becoming non-negligible. This paper aims to compare different methods of classification of every machine learning algorithm on movie subtitle document and then analyse the result of each method. For the classification of movie subtitle document, this paper mostly does the following features of the work: firstly, retrieve of movie subtitle document and pre-processing step of text document. Secondly, deletion of the irrelevant text and extraction of the related texts, the methods are unsupervised machine learning approach and supervised machine learning approach respectively. Furthermore, analyse each classifier in various model which applied with different approach. Lastly, compare the result of different approach and each model then display the results.