Legal Innovation Data Institute was launched Sept. 21
Sept. 21 marked the launch of the Legal Innovation Data institute, a first-of-its kind non-profit aimed at increasing the accessibility of legal data so it can be used to build new legal tools, increase access to justice and enhance personal privacy.
The institute, its partners and collaborators have harnessed a massive volume of Canadian legal information to build the LIDI Data Trust. LIDI hopes providing access to this data will lead to enhanced personal privacy collection; cleaner, normalized and enriched data; the development of free legal apps and the advancement of French language access to justice, the LIDI website states. LIDI will be working with courts and tribunals to continually expand their data sets, says Colin Lachance, LIDI’s founder and executive director.
Lachance says he has wanted to pursue an open-data project such as LIDI “for years.”
“I was of the view that we shouldn't limit innovation to the people who currently have access to the data, that the more hands that bulk data could be in, the more opportunity for innovation,” says Lachance, who was formerly the president and CEO of CanLII and general manager for North America of the global legal publisher vLex.
“It's like the equivalent of paving the road, so that other people can build on it,” says Lachance. “The legal Innovation data Institute is not the one that's going to build the apps. But we're going to make it possible for others to build apps, services, to develop new insights into the evolution of case law.”
Whether they be law firms, corporate legal departments or government legal departments, large institutions are working on improving their knowledge management skills, says Lachance. iManage, HighQ and other services allow them to run their business more efficiently by providing insights from all of their internal documents. LIDI can infuse their existing knowledge management systems with “well-structured, well-tagged and topically tagged data,” he says.
“What that means for them, is they can essentially carry out legal research in the background without having to carry out legal research.”
Access to LIDI’s data will help remedy the problem that traditional legal research methods are not scalable, says Wolfgang Alschner, an associate professor at the University of Ottawa’s Faculty of Law. No lawyer can consume every decision ever rendered by the Supreme Court of Canada, for example, and these limitations prevent lawyers from making truly evidence-based decisions, he says.
“We might pick a small sample of decisions in order to say something about jurisprudential trends at large, or we might look at a couple of decisions by a certain judge in order to predict our chances of being successful in future litigations,” Alschner says.
“Now, once you can scale-up legal analysis, you can make evidence based decisions on a much broader information basis, so you can make more accurate decisions, you can anchor your decisions in empirics, and therefore, make better decisions.”
It is access to the data, not the technology, which holds back the scalability of legal data analysis, says Alschner. Algorithms are cheap and abundant, whereas legal data is neither cheap, nor easy to obtain.
“Because access to legal data was not abundant, the legal analytics ecosystem in Canada was basically non-existent,” Alschner says. “And it was very difficult for researchers, but also for startups to do this type of large-scale, evidence-based text mining.”
By training, Alschner is an international lawyer with an academic interest in legal data science, applying natural language processing, network analysis and machine learning to legal texts, including international treaties. This led him into the “emerging field” of computational legal studies and a collaboration with Lachance, he says.
LIDI’s open-data system will also aid personal privacy for court participants through more effective anonymization, says Lachance. While Canada has an open-court system and the public is entitled to see the details of any court decision, the exposure of the private details of parties or witnesses has made courts reluctant to make data available in bulk, he says.
“If someone is looking at 100,000, Ontario court decisions, and they are using that information to build personal profiles on people mentioned those court decisions, that's a big, legitimate fear,” Lachance says.
But, he adds, “We can now vastly reduce the risk of exposing people's private information.”
A common AI tool which can be used for anonymizing personal information is Named Entity Recognition, with which individuals and organizations can be discerned within vast amounts of caselaw data. But the technology, as it exists now, strips out too much information, says Lachance. LIDI is pursuing a research project with Private AI, a company specialized in the de-identification of data, to develop a caselaw-specific anonymizer. This will allow for versions of documents where private individuals are anonymized and judges, lawyers, police officers, expert witnesses and other public players are not.
Lachance expects LIDI to invite “significant waves of innovation” from players who have not yet been involved in the legal environment. For example, they are partnered with machine-learning company AltaML, a company that builds AI tools for businesses.
Open and accessible data will mean that “legal tech is no longer defined by the giants on one side, like LexisNexis and WestLaw, and entrepreneurs and startups on the other,” says Lachance.
“This opens the door for large scale technology companies to cross industries into legal more easily.”
In a bilingual country such as Canada, another LIDI partnership Lachance says is significant is with BG communications, a translation service. BG wants to advance the science of machine translation in French and through the use of bilingual caselaw, Canadian court data will help them achieve that.
“So the cross-pollination opportunities go way beyond the creation of a new app,” he says.