Analysing the Poetic Structure of Jana-Gaṇa-Mana in Entirety: A Statistical Approach

Measurable investigation of abstract content so as to bring bits of knowledge into its expressive highlights has been a shared zone of enthusiasm among the aficionados of writing and measurements. The thought of applying the equivalent for mining critical perceptions with respect to our huge abstract legacy has likewise intrigued the antiquarians. There have been several works on automated analysis of poetry. Notwithstanding, adequate consideration has not been laid on such works in even Hindi-prose, leaving Hindi-poetry with an almost total nonattendance of any examination in such manner. The curiosity of the work proposed in this paper is emphasized by the way that it includes the utilization of binomial model to literary compositions which means to take a new way towards statistical interpretation of literary pieces. Any poetic composition in Hindi may be viewed as a sequence of numeral symbols 1 and 2 only which correspond to a laghu (short syllable) and a guru (long syllable) respectively. Determining short and long syllables in Hindi-poetry is not a trivial activity as this determination is not independent rather is a function of a number of factors. The candidate poem is statistically analysed using three case studies. In the first study, the laghu and guru data is successfully fitted to a binomial model. Short syllables themselves can be classified into two types. The second case study is an attempt towards fitting a binomial model to the type_1 and type_2 laghu data. In Hindi, a vowel can be written either independently or as a mtr (diacritic mark) along with a consonant letter. As our final case study, we have successfully fitted a binomial model over diacritic marks data in Jana-Gaa-Mana. Importance of this task lies in the observations leading to fit the various phonological features of the candidate song through a single model type, that is, binomial model.