| Abstrakt: |
Hindi is the third largest spoken language in the world and first in India. Addressing research and applications in natural language processing, Hindi, the dominant language in India, is still in its infancy. Hindi title generation is a crucial process that should be taken into consideration for any given story or document. This paper describes the automatic generation of title for a given text in Hindi. We have designed a system which generates the title relevant to the text or story and preserves the central theme. In the proposed system four algorithm are designed after performing the analysis of various text and stories of Hindi. The title will be created by the first algorithm using the utmost priority nouns. Second algorithm will generate the title as the combination of adjective and noun. Third algorithm is based on the proverbs present in the text and the last algorithm is based on the set of keywords present in the text or story. This system first tokenizes the text and then apply the tagging for identifying the nouns, pronouns, adverbs, adjectives etc., and then by applying different designed algorithms suggests the title of the text or story given by the user. Using 10 stories through a well-known Hindi text collection, the proposed approach is tested. Eight of the ten titles it produces are accurate. Second, a group of users, including Hindi teachers and academics, test the system. Their reviews and opinions indicate that the proposed strategy produces findings that are pertinent and is effective. The size of the story affects how well the system works. The most pertinent titles can then be chosen. This approach can be used by Hindi students who are still learning the language, as well as by novelists, newspapers, and magazines to produce titles at random for them. [ABSTRACT FROM AUTHOR] |
