Name transliteration is an area which deals with transliteration of out-of-vocabulary (OOV) words. It plays an important role in developing automatic machine translation and cross lingual information retrieval system because these systems cannot directly translate out-of-vocabulary (OOV) words. In this article, we present SVM based name transliteration approach that considers transliteration task as a multi-class problem of pattern classification, where the input is a source transliteration unit (chunks of source grapheme) and the classes are the distinct transliteration units (chunks of target grapheme) in the target language. Our proposed approach deals with Bengali-to-English forward and backward name transliteration. Our proposed method has also been compared with some existing transliteration model that uses a modified version of Joint-Source channel model. After the systems have been evaluated, the obtained results show that our proposed SVM based model gives the best results among the others.
CITATION STYLE
Sarkar, K., & Chatterjee, S. (2017). Bengali-to-english forward and backward machine transliteration using support vector machines. In Communications in Computer and Information Science (Vol. 776, pp. 552–566). Springer Verlag. https://doi.org/10.1007/978-981-10-6430-2_43
Mendeley helps you to discover research relevant for your work.