Implementation of Smart Legal Assistance System in Accordance with the Indian Penal Code Using Similarity Measures

Abstract

The rate of crime is increasing rapidly and many citizens of country are the victims of these crimes. It is observed that though the number of crimes taking place is very huge, the actual number of crimes being reported to the legal authorities is very small. This huge difference is because of several reasons where one of the reason is lack of awareness about the civil rights and the laws. Many citizens in India are unaware about the evil practices performed against them and unethical exploitation of them done by others. The citizens are also not aware about various laws in the Indian Penal Code. So inorder to bridge this gap between the laws and the common citizens of India, we have proposed a Natural Language Processing (NLP) system using Word Mover’s Distance (WMD) which takes as input the textual description of crime and generates as output, the applicable Indian Penal Code (IPC) sections and the description of the punishments mentioned under these sections. The performance comparison of WMD with cosine similarity is presented. Finally, the accuracy of the system is measured by getting the result generated by the system to be validated by a lawyer personal.