December 18, 2020

 Dr. Daniel Martin Katz

legal data science interdisciplinary teams teams legal informatics social physics network analysis and law legal physics legal networks legal complexity document networks digital transformation nlp legal document analytics legal data artificial intelligence data science technology and digital machine learning law company legal tech

Last month, I co-hosted The Physics of Law Virtual Conference. Overall, we had ~300 Attendees from 40+ Countries watch the presentation of 20 Academic Papers by 62 Authors. We saw a wide range of methods from Physics, Computer Science and Applied Mathematics devoted to the exploration of legal systems and their outputs.

Methodological approaches included Agent-Based Modeling, Game Theory and other Formal Modeling, Dynamics of Acyclic Digraphs, Knowledge Graphs, Entropy of Legal Systems, Temporal Modeling of MultiGraphs, Information Diffusion and Predictive Modeling.  NLP Methods on display included traditional approaches such as TF-IDF, n-grams, named entity recognition and other metadata extraction as well as more advanced methods such as Bert, Word2Vec, GloVe, etc.

Methods were then applied to topics including Attorney Advocacy Networks, Statutory Outputs from Legislatures, the analysis of various bodies of Regulations, Contracts, Patents, Shell Corporations, Common Law Systems, Legal Scholarship and the Regulation of Financial Systems. After undergoing the Peer Review process — Final Papers to be published in Frontiers in Physics in 2021.

While of course, not every idea is useful or relevant to the commercial sphere, conferences such as this are where many ideas (some of which ultimately land in commercial products) are first proposed or displayed. Sometimes these developments happen right away and sometimes it requires a number of years of refinement before they come to market. For those who are interested in other related conferences – see e.g. the Workshop on Natural Legal Language Processing (NLLP), Jurix and IAAIL and many others!

I would encourage folks to continue to monitor fields such as Physics, Computer Science (ML, NLP), Applied Mathematics, etc. as it is these fields (among others) that offer many ideas that can power future legal technology offerings.  Of course, these ideas alone are insufficient as they must still be projected in the specific work that we do here in the legal sector.

So how does one actually undertake this translation exercise? That translation occurs when a team can truly span both intellectual spheres (law & tech).  In other words, the key is to construct a team with demonstrable/verifiable technical expertise not just within a given domain (law or tech) but across the domains (law and tech).

Here is the key point as I see it – it is only through interdisciplinary collaboration that will we truly solve some of our most challenging problems in law.  A few years ago, I called it {Law, Tech, Design, Delivery}.  It manifests today in the idea of #LawCompany.  The idea of a Law Company is to be broader than a single product or service. The idea of a Law Company is about offering more than just a collection of substantive legal experts.  Instead, a Law Company is a place where multidisciplinary teams work to solve everyday (BAU) work in a cost-effective manner.  A Law Company is a place where ‘best in class’ methods can be applied to some of the most complex problems in the field.