RETRACTED ARTICLE: Sentimental analysis of transliterated text in Malayalam using recurrent neural networks

Abstract

Usage of mobile phones, access to internet in the fingertips and increasing number of mobile applications has accelerated the generation of online content. Freedom of expression has waived the barriers of online interaction. Curiosity in knowing others viewpoint through their reviews in each and everything starting from a purchase of product to watching a movie has become a common scenario. Decision on success and failure of one’s business is in the hands of public now. Humans are always comfortable, sticking to their native Language, when it comes to expressions whether it is interest, emotions, feeling or opinion. Usage of Natural Language and the trend to analyze the subjective sentiments is increasing day by day. Transliterated text of a language is the English version of spoken native language. For example, Malayalam in English we call as Manglish. Transliterated text has become the language of social media websites like WhatsApp, Facebook, Twitter. It’s a kind of boon to the young generation who know to speak their native language but not to read or write in its own nominal scripts. In this paper we consider the Sentimental Analysis of Transliterated text. RNN-LSTM technique is used to derive the sentiments of transliterated text.