Computational linguistic prosody rule-based unified technique for automatic metadata generation for Hindi poetry

Metadata generation for the poems based on the unified rules is very complex from the viewpoint of computational linguistics. This is more tedious when it comes to Hindi poems. Prosody or ‘Chhand’ as it is called in the Hindi language consists of several sets of rules and which are used while construction of a Hindi poem. Currently, no such metadata generator or technique is in existence which can generate the metadata of Hindi poetry based on the prosody or any other Hindi grammatical rule. In this research paper, we are trying to wrap up all the difficulties with respect to the metadata generator based on ‘Chhands’, Types of ‘Chhands’, Parts of ‘Chhands’, Classification of ‘Chhands’. Along with this paper majorly focuses on the unified-rule based technique for the generation of metadata based on the different set of rules of prosody. On 3026 different UTF-8 based inputs of poems, part of poems, stories, etc., We were able to achieve 98.09% accuracy with the implementation of this unified-rule based technique of metadata generation, rest 1.91% failure was mostly due to input data related issue.