Caselaw Access Project

2018: Project Implementation

At the end of October, 2018 the Library of School of the right of Harvard University (Harvard Law School Library) completed  the project on digitization lasting five years more than 40 million documents in which real legal cases are described. They were uploaded publicly for the purpose of training of computers of law.

As notes the MIT Technology Review edition, one of the biggest problems on the way of development of artificial intelligence for legal applications is the lack of access to sufficient amount of data. For training of such algorithms in their developers often it was necessary to create own bases, filling them with poor data from public websites, or to agree with the companies for gaining access to their legal files.

The Harvard digitized millions of legal cases for training of AI

Thanks to the project of Harvard which received the name Caselaw Access Project millions of real legal cases became available on the Internet to all of persons interested. Thus, the big source is provided for legal information which can be used for a training of robots lawyers.

Besides, programs will be able easily to look for the necessary text and to provide to "living" lawyers necessary information for work on this or that case.

The managing director of Library Innovation Lab laboratory at Library of School of the right of Harvard University Adam Ziegler considers that thanks to Caselaw Access Project there will be a set of experiments, and progress in legal training of artificial intelligence will accelerate.^[1]

It is really very difficult to construct the smart interface if you cannot reach basic data — he noted.

The materials which are laid out by Harvard include the judicial conclusions and shorthand: they contain complete available materials on each legal case. The first documents are dated 1658, and the last — the 2018th.

Notes