For Immediate Release:  Jan 11, 2012
2012, 01, 11

Drawing on Science to Advance the Rule of Law

Drawing on Science to Advance the Rule of Law

Media Contact

Kristen McMahon
Director of Public Relations
Maurice A. Deane School of Law at Hofstra University
Phone: 516.463.4252
E-mail: Kristen.D.McMahon@hofstra.edu

HEMPSTEAD, N.Y. — A scholarly article featuring research conducted by the Research Laboratory for Law, Logic and Technology (LLT Lab) at the Maurice A. Deane School of Law at Hofstra University has been published in the peer-reviewed journal Artificial Intelligence and Law. The article, “A Framework for the Extraction and Modeling of Fact-Finding Reasoning from Legal Decisions: Lessons from the Vaccine/Injury Project Corpus,” was co-authored by Vern R. Walker, professor of law and director of the LLT Lab, and former student researchers Nathaniel Carie ’11, Courtney C. DeWitt ’11 and Eric Lesh ’11.

"In an era of automated question-answering, such as IBM’s Watson of Jeopardy! fame, it will become increasingly important to develop and annotate text corpuses from human decision-makers in terms of models of expert decision-making,” said Kevin D. Ashley, editor-in-chief of Artificial Intelligence and Law. “That is the kind of serious empirical work Professor Vern Walker and his students have undertaken with the Vaccine/Injury Project Corpus of legal decisions and accompanying models of fact-finding.”

The article highlights the methodologies being developed and employed by researchers working in the LLT Lab’s Vaccine/Injury Project Corpus. The corpus consists of legal decisions awarding or denying compensation for health problems allegedly due to vaccinations, and the accompanying logic models created by the researchers. It thereby provides useful data for logic theory, natural-language linguistics and artificial intelligence research. 

The article also recounts the lessons learned by the LLT Lab researchers as they refined protocols for the manual extraction of logic structure from legal decisions to generate logic models. By incorporating examples drawn from the corpus and sharing insights extrapolated from their research, the authors shed some light on how this data-extraction process might be replicated through automation.  

“Models like these will help AI researchers determine how automated question-answerers can reach and explain their decisions based on rule-trees and other argument schema gleaned automatically from texts,” Ashley added. “Reasoning models and annotated corpuses like Walker’s will jump-start the machine learning and information extraction needed to finish the task. The fact that law students are gaining practical legal knowledge and expertise as they perform these annotations invites the American legal academy to take note.”

“My experience in the lab has definitely helped me develop evidence-assessment and deductive-reasoning skills,” added Nathaniel Carie ‘11. “It is exciting to be able to combine legal research and advanced software to develop tools that will have a significant impact on the way we think and practice the law of the future.”

In an effort to become more efficient, many law firms are relying on software — for example, electronic discovery programs — that mimic many of the human processes involved in tasks that were once exclusively the province of lawyers. The goal of the research at the LLT Lab is to develop accurate and cost-effective methods for extracting the logic structure from legal decisions that can be standardized and applied in different legal contexts for use by judges, lawyers and educators to increase transparency, reliability and access to justice.  “Our goal is not only improving legal research and education, but also having an impact on legal decision-making in society,” said Walker. “In this article, we reflect on our research and education methods in order to help improve computer software for argumentation mining.”

Information about the LLT Lab, including its mission, methodologies and work product, can be found at LLTLab.org

# # #

The Maurice A. Deane School of Law at Hofstra University is located 40 minutes from New York City in suburban Long Island. Hofstra Law is home to nearly 850 students, an alumni base of more than 10,600 members and a distinguished faculty of more than 50 professors, including many scholars recognized as national and international experts in their field. The law school is part of Hofstra University and is fully accredited by the Council of the Section of Legal Education and Admissions to the Bar of the American Bar Association.

Hofstra Law

121 Hofstra University

Hempstead, NY 11549-1210

Map | Directions

Phone Information

tel: 516-463-5858

fax: 516-463-1017

Directory

Email Information

hofstralaw@hofstra.edu

Directory