Problem Statement- Given the facts / description of a case, predict the legal articles violated in the given case. [Dataset- ECtHR (European Court of Human Rights)]

Ongoing Work:

1. Heterogeneous Graphs Existing LJP techniques use a single embedding for large case documents [Chalkidis et al. (2019)]. Compute fine-grained embeddings with different kinds of nodes- facts, cases, articles

2. Summarization Query LLMs to summarize an entire case or parts of it. Could learn better case embeddings and improve the performance of the GNN models.

References:

1. Cui, J., Shen, X., & Wen, S. (2023). A survey on legal judgment prediction: Datasets, metrics, models and challenges. IEEE Access.
2. Chalkidis, I., Fergadiotis, M., Malakasiotis, P., Aletras, N., & Androutsopoulos, I. (2020). LEGAL-BERT: The muppets straight out of law school. arXiv preprint arXiv:2010.02559.
3. Veličković, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., & Bengio, Y. (2017). Graph attention networks. arXiv preprint arXiv:1710.10903.
4. Chalkidis, I., Androutsopoulos, I., & Aletras, N. (2019). Neural legal judgment prediction in English. arXiv preprint arXiv:1906.02059.
5. Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., ... & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33, 9459-9474.
6. Chalkidis, I., Fergadiotis, M., Tsarapatsanis, D., Aletras, N., Androutsopoulos, I., & Malakasiotis, P. (2021). Paragraph-level rationale extraction through regularization: A case study on European court of human rights cases. arXiv preprint arXiv:2103.13084.