SUM to develop a legal document analysis system based on neural networks
A pilot version of the software is planned to be presented by mid 2020.
Researchers at the State University of Management (SUM) are working on the creation of an intelligent legal document search system ‘Yurtech’ (‘LegalTech’), which will use neural networks to find contradictions in laws and suggest which documents will be modified by the approved amendment, SUM Rector Ivan Lobanov told TASS on Wednesday.
Researchers from the State University of Management have designed a digital solution for legislative authorities and lawyers to analyse amendments and by-laws for contradictions and semantic links with the current legislation of different levels. This system uses neural networks to create semantic selections, rather than merely selecting texts according to given keywords. Compared to existing legal databases, it will enable a more accurate analysis and wide coverage of data,
he explained.
The intelligent search system provides the ability to download full texts of legislative documents for analysis, thereby improving the quality of search results. The neural networks select keywords independently.
The designed semantic analysis algorithms are currently undergoing tests by staff at the SUM Institute of Public Administration and Law. By the middle of 2020, the developers intend to present a pilot version for implementation at the State University of Management,
added Lobanov.
Подпишись на тг-канал "Наш ГУУ"
Дата публикации: 13.11.2019