Column

AI4A2J

Lexum
Author: Ivan Mokanov, Lexum

The title of this post stands for Artificial Intelligence for Access to Justice. It sounds a little like buzzword festival. Rest assured however – there is no mention whatsoever of block chain or design thinking further down in the text.

A few months ago we sent out an invitation to industry partners to join Lexum Lab (Lexum’s R&D team) to test a few AI / Deep Learning applications that are in the making. More specifically, Lexum Lab and the Montreal Institute for Learning Algorithms (MILA) are collaborating on the development of a link prediction algorithm for law. Link prediction attempts to solve the following problem: given a snapshot of a network, can one predict the next most likely links to form in the network? You can see prediction algorithms in popular platforms today. This is how Facebook suggests friendships and how Tinder proposes matches.

Our intended use of link prediction is slightly less romantic. Based on a large network of legal citations (12.8M links to cases and 13.4M links to statute sections), our goal is to create suggested links between legal documents that didn’t exist before. We published more details about our initial take on this project here.

Lexum will be collaborating with the Canadian Bar Association’s Immigration Section on a project geared towards improving access to justice in the field of immigration law.

The Federal Court of Canada’s Bench and Bar (Immigration Liaison) Committee has recently formed a group of pro bono lawyers who will assist self-represented litigants appearing before the Federal Court of Canada on immigration matters. The initiative is very new and the scope of the group’s work is still being defined, but it may include general assistance with the case, procedural guidance, settlement aspects, and presentation of possible outcomes of cases. In order to support the members of this group, a web application based on Lexum’s link prediction technology will suggest relevant federal court cases when presented with the Immigration and Refugee Board (IRB) decision submitted for judicial review by the Federal Court of Canada. In other words, Lexum’s solution will pinpoint what the Federal court decided in the past on matters similar to the ones the self-represented litigant is dealing with in the present.

Putting our technology at use to assist pro se litigants while getting real-world feedback on our Lab work makes this project a mutually beneficial collaboration for all parties involved. It also spurs ideas about potential applications in other fields of law facing high volumes of cases involving largely unrepresented parties and having important social implications.

Many thanks go to Marina Sedai, National Chair, CBA Immigration Section, Michael Battista, Chair of the Bench and Bar subcommittee and Andrew Baumberg, Legal Counsel at the Federal Court of Canada for their vision and action in this worthwhile initiative.

Start the discussion!

Leave a Reply

(Your email address will not be published or distributed)