This repository consists of all NLP related experiments conducted on Indian Legal-Judicial data by the JusticeHub team.
We selected a sample of 51 POCSO judgements from Assam and checked how OpenNyAI can be used to extract variables from the judgements to perform legal research. This repository has codebooks of this experiment. The inferences frm this experiment are shared in this blog: Link
The experiment was presented at a HydPy meetup. The talk can be accessed here:Video and the slides used for the talk are here: Google Slides
NERs are useful for recognising "Entities" in a piece of text. For Court Judgements, these entities could be "STATUTES", "JUDGE NAME" etc. This directory has codebook that experimented various NER Models on Indian judicial texts.
Summarization Models help in summarizing a piece of text. This directory has codebooks on extractive summarization of Indian court judgements using OpenNyAI library.