AUTOMATIC DENTIFICATION SYSTEM FOR POETIC DEVICES AND THEMATIC PATTERNS IN HINDI LITERARY USING NLP

Abstract

The present invention relates to an automatic identification system for detecting poetic devices and thematic patterns in Hindi literary texts utilizing Natural Language Processing (NLP) techniques. This system is designed to analyze both structural and semantic elements of Hindi poetry and prose by implementing a multi-stage process that involves the following key components: 1. Text Preprocessing: Includes tokenization, lemmatization, and normalization adapted specifically for Hindi linguistic structures. Custom algorithms handle the complexities of compound words, postpositions, and inflections in Hindi. 2. Poetic Device Detection Module: This module leverages rule-based methods and machine learning classifiers to detect poetic devices. These include: o Metaphor and Simile Detection: Using semantic similarity algorithms and dependency parsing to identifY figurative language. o Alliteration and Rhyme Scheme Identification: Phonetic analysis tools are developed to detect sound-based poetic devices.o Meter and Rhythmic Patterns: A prosody analysis submodule detects traditional Hindi poetic meters like Chhand using syllable segmentation and pattern matching techniques. 3. Thematic Pattern Recognition Module: This component uses topic modeling (e.g., · Latent Dirichlet Allocation) and sentiment analysis to extract recurring themes and emotional tones within texts. The module also employs Named Entity Recognition (NER) tailored for Hindi to identify cultural, historical, and mythological references that are prevalent in Hindi literature. 4. Deep Learning Integration: A deep learning model (e.g., Transformer-based models like BERT or GPT fine-tuned for Hindi) is employed to improve the system's accuracy in detecting complex linguistic structures and nuanced themes. These models are pretrained on extensive Hindi literary corpora and fine-tuned for specific poetic devices and thematic patterns. 5. Explainability and User Interface: The system includes an explainability component that provides transparent .results to users, highlighting which textual features led to the identification of specific poetic devices or themes. The user interfuce is equipped with visualizations that map the occurrence of devices across the text and offer insights into thematic shifts.

Documents