Resolving Anaphora using Gender and Number Agreement in Marathi text
Kalpana B. Khandale, C. Namrata Mahender · 2021
In the area of natural language processing, there is a wide variety of text data available, but not in the standard format and therefore it is necessary to standardize it. This paper concentrates on an anaphora resolution for Marathi text because the anaphora has been resolved in various foreign as well as Indian languages like Hindi, Bengali, Kannada and many more but very less work has been done in Marathi. Anaphora often occurs written text and spoken dialogue and it is necessary to resolve an anaphora to better understanding of the discourse or the sentence because the a human being we can easily find out the anaphora refer for what antecedents but for machine it is very difficult task and hence resolving anaphora becomes a most challenging task for the researcher Here this paper concentrate to resolving the anaphora in Marathi text using the gender and the number agreement as well as the animistic knowledge. And the overall performance achieved by the system is 68.88%.