First Artificial Intelligence Data Set Created for International Law Texts

The data set in AI studies plays a critical role to learning an algorithm and giving successful results. IntLawNER, researchers presented an auditory data set called IntLawNER in order to resolve this disclosure. This data set aims to develop systems that provide special names and concepts in law texts.

Basic features of IntLawNER data set:

• 2.987 sentences were manually processed from International Court of Justice (ICJ) decisions
• Total 8.094 individually named asset (entity) span labeled by human experts Tags
• Other international law sources such as UN Security Council texts Tags
• It is the first academic data set structured at token level
• Open sourced in the format that researchers can use immediately

International law manages critical areas such as coordination of states, regulation of armed conflicts and human rights protection. However, it was necessary to develop artificial intelligence tools that are specific to this field, quality and open data sources have been missing so far. IntLawNER fills this gap and opens the door to the development of the legal AI applications.

Such data sets will be able to automatic processing of systems, contract analysis tools and international agreements that will help legal students. It was a reliable foundation for new studies to be done in technology and legal intersection.