Human languages are handled in different ways at different levels, such assyntactic analysis at sentence level, semantic analysis at meaning level, discourse analysis at text level and morphological analysis at word level. Syntactic analysis or parsing is an important application in field of Natural Language Processing (NLP) as it helps in determining underlying meaning and depicting the output as linking of sentences to each other. Parsing is performed to determine the grammatical structure, marking the partsofspeech and specifically to remove ambiguity. Though there is no way to remove ambiguity completely, parsers partially remove it. There are various types of ambiguity and for each type; there are different ways or methods to handle. This paper presents the different approaches followed for parsing across different languages such as Chinese, Arabic, Mongolian and Hindi. Some of them followed the static approach while some followed dynamic approach. In case of machine learning models, some of these languages followed supervisedlearning model while others followed unsupervised model. Their comparisons have been laidto create a parser or word sense classifier or translator or Finite State Automata(FSA) parser, each aiming to achieve maximum accuracy and optimum results.
展开▼
