Comprehensive analysis for assessing the effectiveness in the implementation of word sense disambiguation in the Gujarati language

Natural language is ambiguous, whereby numerous words have different interpretations contingent upon the surrounding context. Ambiguities in language pose significant challenges, as words in human language can have multiple interpretations depending on the context. Word sense disambiguation is critical and substantial in the initial stage of natural language processing. It refers to the computational ability to determine the specific meaning of words within a given context. It is the process of determining the correct meaning of a word from a list of potential meanings, relying on the surrounding context in which the word is used. This article makes multifold contributions. First, a comprehensive survey explores various methodologies used in different research studies and the current state of performance in the WSD field. Second, this paper describes the supervised, unsupervised, hybrid, and knowledge-based WSD approaches explored for the Gujarati language. Third, the availability of various wordnets and datasets is presented. Fourth, different machine learning algorithms are implemented using a self-generated annotated sense dataset to assess the WSD in the Gujarati language. Fifth, existing challenges in WSD approaches and future directions are presented. Overall, this study helps the researchers work in machine translation, lexicography, question answering, information retrieval, word processing, text summarization, and text classification.