Advancing Legal Decision Making: A Unique Method for Classifying Offense Incidents under the Indian Penal Code using Word Embedding and Labeled Searching

The advancements in artificial intelligence (AI) have propelled the domain of text similarity, where sophisticated algorithms harness the power of natural language processing to analyze and compare textual content. In the legal domain, the judicial system constitutes a vast landscape generating copious amounts of daily textual data, encompassing an array of documents such as court records, investigative reports, and First Information Reports (FIR). Moreover, a substantial amount of written content within this domain is found in legal articles and regulatory manuals. The similarity between these texts aids us in some way in making decisions. The Indian Penal Code (IPC) serves as the principal legislative framework in India for delineating criminal offenses. In the context of computer programs, maintaining information in a well-organized and structured format is crucial. This underscores the need for a suitable knowledge base that accurately represents the sections of the Indian Penal Code. Through evaluating various word embedding techniques and cosine similarity for measuring text similarity between crime complaint report and offense terms, this research reveals a method for categorizing IPC Sections. In addition, this study seeks to present a novel approach for creating a specialized knowledge base tailored to IPC documents. It also aims to introduce a labeled searching technique to augment the precision and efficacy of the acquired results.