Natural language processing for Nepali text: a review

Tej Bahadur Shahi, Chiranjibi Sitaula

Research output: Contribution to journalArticleOtherpeer-review

12 Citations (Scopus)

Abstract

Because of the proliferation of Nepali textual documents online, researchers in Nepal and overseas have started working towards its automated analysis for quick inferences, using different machine learning (ML) algorithms, ranging from traditional ML-based algorithms to recent deep learning (DL)-based algorithms. However, researchers are still unaware about the recent trends of NLP research direction in the Nepali language. In this paper, we survey different natural language processing (NLP) research works with associated resources in Nepali language. Furthermore, we organize the NLP approaches, techniques, and application tasks used in the Nepali language processing using the comprehensive taxonomy for each of them. Finally, we discuss and analyze based on such assimilated information for further improvement in NLP research works in the Nepali language. Our thorough survey bestows the detailed backgrounds and motivations to researchers, which not only opens up new potential avenues but also ushers towards further progress of NLP research works in the Nepali language.

Original languageEnglish
Pages (from-to)3401–3429
Number of pages29
JournalArtificial Intelligence Review
Volume55
DOIs
Publication statusPublished - 27 Oct 2021

Keywords

  • Classification
  • Devanagari
  • Machine learning
  • Natural language processing
  • Nepali language
  • Nepali linguistics
  • Sentiment analysis

Cite this