Computational Analysis of Figurative Language for Koshur and Awadhi Poetry

Poetry is a distinct literary form that creates powerful emotions, breaks down difficult ideas into digestible parts, and conjures up powerful imagery through the use of language and imaginative techniques. Rhyming schemes, rhythmic meter, and metaphorical language are common devices used in this spoken or written art form to captivate listeners on multiple levels. Poetry is an expressive form of literature, but up until recently, figures of speech (FsoS) in poetry could not be automatically detected by natural language processing (NLP) approaches. In this study, authors have designed four algorithms for figures of speech like प्रश्ना (‘Prashna,’ ‘Question Framing’) and लाटानुप्रास (‘Lata,’ ‘Word Repetition’) for two languages Koshur and Awadhi. The annotation, computational analysis, and comparison of six FsoS notably व्रत्या अनुप्रास (‘Vratya,’ ‘Type of alliteration’), छेकानुप्रास (‘Checka,’ ‘Type of alliteration’), अन्तत्यानुप्रास (‘Anantya,’ ‘Rhyme’), पुनरुक्ति (‘Punurukti,’ ‘Repetition’) including प्रश्ना (‘Prashna,’ ‘Question Framing’), and लाटानुप्रास (‘Lata,’ ‘Word Repetition’) are performed. Above 90 and 87% result is achieved during the computational analysis for Koshur and Awadhi languages, respectively.