A survey on novel approach to semantic computing for domain specific multi-lingual man-machine interaction

Abstract

Natural language processing (NLP) helps computational linguists to understand, process, and extract information from natural languages. Linguist Panini signifies ’information coding’ in a language and explains that Karakas are semanticosyntactic relations between nouns and verbs that resemble participant roles of modern case grammar. Computational grammar maps vibhakti (inflections) of nominals and verbs to their participant roles. Karaka’s theory extracts semantic roles in the sentences which act as intermediate steps for various NLP tasks. The survey shows that NLP seeks to bridge the gap for man-machine interaction. The work presents the impact of machine learning on natural language processing with changing trends from traditional to modern scenarios with Panini’s classification scheme for semantic computing facilitating machine understanding. The study presents the significance of Karaka for semantic computing, methodologies for extracting semantic roles, and analysis of various deep learning-based language processing systems for applications like question answering. The survey covered around 50 research articles and 21 Karaka-based NLP systems performing multiple tasks like machine translation, question-answering systems, and text summaries using machine learning tools and frameworks. The work includes surveys from renowned journals, books, and relevant conferences, as well as descriptions of the latest trends and technologies in the machine learning domain.