The AIIR Lab (Artificial Intelligence and Information Retrieval Lab) at the University of Southern Maine (USM)recently participated in the 12th Competition on Legal Information Extraction and Entailment (COLIEE-2025), an internationally recognized challenge focused on advancing the field of legal AI. Graduate students Deiby Wu and Sarah Lawrence, both alumni of USM’s Computer Science undergraduate program, represented the lab in this prestigious event. Sarah is currently enrolled in USM’s 4+1 accelerated master’s program, which enables students to complete both their bachelor’s and master’s degrees in just five years.
COLIEE-2025, a global benchmark competition, is designed to foster research in legal text processing, with tasks that test systems’ capabilities in legal case retrieval, legal entailment, and statute law understanding. The competition includes four primary tasks:
- Task 1: Legal Case Retrieval
- Task 2: Legal Case Entailment
- Task 3: Statute Law Retrieval
- Task 4: Legal Textual Entailment
Deiby and Sarah worked collaboratively on all four tasks, developing methods that integrate cutting-edge large language models (LLMs) with traditional legal information retrieval strategies. Their team’s submission was accepted for presentation at the International Conference on Artificial Intelligence and Law (ICAIL) 2025), the premier conference in the field of AI and law. Their paper, titled “AIIR Lab at COLIEE 2025: Exploring Applications of Large Language Models for Legal Text Retrieval and Entailment”, marks both students’ first peer-reviewed academic publication.
ICAIL, held biennially and sponsored by the International Association for Artificial Intelligence and Law (IAAIL), brings together leading scholars and practitioners working at the intersection of computer science, law, and ethics. The 2025 conference featured keynote talks, workshops, and technical sessions covering the latest advances in legal reasoning, natural language processing (NLP), and automated legal decision support systems.
Their participation in COLIEE and presentation at ICAIL provided both students with valuable experience applying AI and machine learning techniques to complex, real-world legal data. This work reflects the growing strength of the USM Computer Science Department’s graduate programs, including the 4+1 accelerated master’s program, which allows talented students like Sarah to begin graduate-level work while completing their undergraduate degrees.
The project also demonstrates the mission of the AIIR Lab to train students in responsible, cutting-edge artificial intelligence research with practical applications. As they continue their graduate studies, Deiby and Sarah look forward to building on this foundational work and exploring further opportunities to contribute to the field of AI for legal and information systems.