Tutorial Schedule

Introduction to Legal Text Analytics (20 mins)

Vasudha Bhatnagar

We will introduce the challenges in Legal Text Analytics and highlight the main ongoing initiatives in this field as well as the research efforts done by the Indian NLP community.

Legal Knowledge Graphs (40 mins)

Balaji Ganesan

The distillation of knowledge from a legal judgment that typically has an irregular structure is essential for any legal text analytics (LTA) task. Most LTA tasks require extraction of semantics-rich information from a legal text that includes concepts, entities, and their relationships. We will discuss various approaches for constructing legal knowledge graphs tailored for Indian legal domain to make the knowledge queryable.

Automatic Question Answering (40 mins):

Muhammad Abulaish

Question Answering (QA) is an effective way to extract user-specific information from a huge amount of textual data. Guided QA is particularly advantageous for legal domain, since judgements contains a huge amount of textual data representing both laws and their interpretations, mainly in unstructured format. We will discuss methods for the design and development of guided-QA system from the legal judgements.

Legal Text Classification (40 mins)

Vikas Kumar

The multi-class classification of legal text into classes defined by legal experts leverages several downstream tasks such as judgement summarization, extraction of law points, question answering and construction of legal knowledge graphs. In this session, we will discuss approaches for the assignment of labels to legal text.

Judgement Summarization (40 mins)

Vasudha Bhatnagar

Legal document summarization entails unique challenges due to the extra-long size. In Indian context, the lack of a standard way of writing judgements, exacerbates the difficulty of the task. We cover prominent generic extractive and abstractive summarization methods, followed by recently proposed algorithms for summarization of legal judgements.