期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：2
页码：1908
DOI：10.15680/IJIRCCE.2017.0502153
出版社：S&S Publications
摘要：Transliteration becomes quite challenging when it comes to convert texts such as names and technicalterms from one script to another using approximate phonetic or spelling equivalent of the other language. In this paperwe are addressing development of a system to recognise the lyrics of songs and displaying in either English or Hindiafter transliterating into the user preferred language. The proposed system uses Audio Recognition as a part of NaturalLanguage Processing for the input and further extraction of the sound inventories of the lyrics and their conversion inrequired form using generative model in several stages. As a detail description to the music media running, briefinformation is displayed such as album genre, artist, composer and other attributes to the playing song. We describeand evaluate the process of transliteration of songs from English to Hindi or vice versa for complete understanding ofpronunciation of the words in preferred language.
关键词：Transliteration; Natural Language Processing; Information Extraction; Machine Translation
