• ENJULA UCHOI
• LENIN LAITONJAM

: Considering the increasing uses of multilingual text, the need for an automatic word level language identification model is raised. In this regard, we present an unsupervised model for word level language identification of English and Kokborok code-mixed and code-switched sentences. Several works have already been reported for various languages, including various Indian languages. But, to the best of our knowledge, ours is the first language identification work dedicated to low resource English-Kokborok language pairs. The proposed model combines a frequency lexicon based, character n-gram language model and a language dependent morphological dictionary-based model for correctly classifying each word. The model which is suitable for low resource languages that do not have a large number of annotated dataset is able to achieve a good performance with word accuracy level of 84%.

Word level language identification, code-mixing, code-switching, dictionaries, affixes, kokborok, English.

"An Unsupervised Word Level Language Identification of English and Kokborok Code-Mixed and Code-Switched Sentences ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 8, page no.2151-2157, August-2020, Available :http://www.jetir.org/papers/JETIR2008286.pdf

"An Unsupervised Word Level Language Identification of English and Kokborok Code-Mixed and Code-Switched Sentences ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 8, page no. pp2151-2157, August-2020, Available at : http://www.jetir.org/papers/JETIR2008286.pdf