Smart Legal Assistance Chatbot for the Pakistan Penal Code

This project introduces an innovative legal chatbot
based on AI that serves users within the consultations concerning
the Pakistan Penal Code (PPC), and it exploits the advanced
automatic learning methods (ML) and natural language understanding (NLP), i.e., cosine similarity, latent semantic analysis
(LSA) and dialogue. These tickets are interpreted by the system
to provide a stable legal guide, which is taken out of a thoroughly
cleaned, formatted data set derived by extracting the contents
of ”PPC.pdf” by use of gift and analysis. In the investigation,
the problem of the changing role of NLP and AI in the area
of criminal justice and civil administration under consideration
is brought to the fore, which includes punishment prediction
models based on crime and intelligent assistance systems in the
field of law that enhance decision making, fairness, production,
and efficiency across the range. Chatbot cosine similarity in
combination with TF-IDF is made to overlap a pertinent part of
PPC to retrieve facts, whereas LSA is made to raise the level
of semantic interpretation and di-alogpt, useful to engage in
aesthetically appealing dialogue on general legal topics. These
NLP modules are incorporated into a structured legal database
via a bottle backend, but also with the possibility of a simple
front on modern web technologies in order to interact with it.
The PPC data set has clear headings of section and titles numbers
outlining crimes, sanctions and sub sections/ clauses to have a
nuance of ideas in conditions and aspects