Abstract
This paper embodies the design of parsing algorithms tangibly for a Bengali parser. To design parsing algorithms a detailed study on linguistics and grammar has been performed. A detailed study also has been made on the various techniques and algorithms of the parsers which have been designed for various languages such as English, Arabic, Hindi, Warlpiri(an Australian language) etc. Finally, Bengali sentences have been analyzed according to Chomskyan grammar. Bottom-up parsing technique and Context-sensitive rules have been used to design parsing algorithms. The designed parser can parse simple declarative sentences for all types of tense i.e., Present, Past, Future and Habitual past tense together with all types of person i.e., First, Second and Third person. The algorithms can maintain the agreement of Person + Class between Verb form and Subject which is available in Bengali sentences. Words are stored into the dictionary (lexicon) together with lexical categories. Using these algorithms it would be possible to parse the Bengali sentences to let the users know whether those sentences are syntactically correct or not.