AILA 2021: Shared task on Artificial Intelligence for Legal Assistance

AILA 2021 was the third edition of the Shared task on Artificial Intelligence for Legal Assistance, that was organized with the FIRE 2021 conference. This year two tasks were offered. While the rhetorical role labelling task was continued from last year, a new Legal Judgement Summarization task was introduced in the current edition. In the Rhetorical Role Labelling task, given a case document, the sentences in the document were to be classified into 7 rhetorical roles – Fact, Ruling by Lower Court, Argument, Precedent, Statute, Ratio of the decision and Ruling by Present Court. The legal judgement summarization task consisted of two subtasks. Subtask (a) was a binary classification task that required participants to identify ‘summary-worthy’ sentences in a court judgement. For subtask (b) participants had to automatically generate a summary from a given court judgement. Datasets for both tasks were created by annotating publicly available judgments from the Supreme court of India.