Generation and Splitting of the Compound Words in Nepali Text

In Nepali language, compound word formation is mostly associated with inflection, derivation, and postposition attachment. Inflection occurs due to suffixation, whereas derivation driven by both prefixation suffixation. The generated the rules may produce lots of out-of-vocabulary words limited lexical resources numerous exceptions. Hence, machine learning approach can help generate valid compounds split them into morphemes that be further used as a resource for spelling suggestions, information retrieval, translation. this research, method from corresponding splits (head prefix/suffix/ postpositions) suggested. A BiLSTM based deep was words. Publicly available Brihat Shabdakosh data Nepal Academy scraped news were experimentation. obtained results found outstanding compared rule-based applied similar job.