AI and data science will help the judiciary protect the environment

Strengthening the role of the judiciary in protecting the environment using artificial intelligence. who – which It is one of the project objectives that have been achieved By the United Nations Development Program (UNDP), the National Council for Justice (CNJ) and the Federal University of Rio Grande do Norte (UFRN). The initiative will use artificial intelligence and data science techniques to extract useful information from procedural texts in order to make analyzes and forecasts in litigation related to the topic of environmental law.

As Raphael Light, Assistant Judge at the Presidency of the National Council of Journalists explains, the project aims to use artificial intelligence, natural language processing techniques, process data analysis and the textual content of its clips to obtain information about the judiciary’s environmental performance. “Artificial intelligence is an important tool to support the work of judges and increase the effectiveness of the judiciary – For example, identifying patterns of behavior, the impact on specific biomes, and the cumulative effects of certain activities –as well as better directing inspection procedures to combat illegal deforestation and other environmental crimes, the judge says.

CNJ already monitors lawsuits in environmental law matters through SireneJud . interactive board, which collects information, for example, about indigenous lands and areas of deforestation. The panel consumes data from various sources, such as DataJud – the national database of the judiciary, the National Institute for Space Research (Inpe) and the Brazilian Institute of Geography and Statistics (IBGE).

UFRN will allow SireneJud APIs (APIs) was developed to add new information to the interactive board. APIs allow systems to communicate and integrate data.

general engineering

In the project, the specific architecture for developing solutions is part of computational linguistics, which uses neural networks and other methods to “extract meaning from text,” explains Elias Jacob, professor at UFRN. “Artificial intelligence is nothing more than a representation of the world from the data passed to it. What does it do? It detects patterns,” describes Jacob.

The Codex platform It will provide the dataset that analysts and developers will use to build AI models and other solutions. “The Codex Alimentarius imports information from the various electronic procedural systems used by courts. It extracts both process metadata, such as the number and name of parts, as well as the content of existing texts,” explains the professor.

according to Codex publication control panel, the system extracted information from 141 data sources related to the procedural systems of Brazilian courts. “The whole project has the function of detecting patterns to extract information – previously inaccessible – so that a human can make the necessary decisions,” Jacob says.

The use of artificial intelligence in the judiciary

Elias Jacob also suggests demystifying the idea that AI models will promote some kind of robotics in judges’ work. “What should be clear is that they are tools that help us do something, not that they came to replace human labor,” he says.

He points out that the social demand for the service provided by the judiciary is high, and even with the digitization of the judicial process and the high productivity of judges and servants, the judiciary faces problems of procedural overcrowding. In this sense, artificial intelligence it’s away: “In Brazil, judges judge more than 9 thousand cases a year. There are several ways to better meet this demand. One of them is the development of artificial intelligence tools.”

The project is being implemented with UFRN under the Justice 4.0 program. The initiative seeks to strengthen the role of the judiciary in protecting the environment, given that Brazilian justice has a body of relevant information and data on environmental disputes and crimes, Jacob explains. “The problems of society lie with the judiciary which, in theory, It is the best source of knowledge about his problems country; On the issue of the environment, it would be no different, ”the professor argues.

Discover the products provided in the partnership between CNJ, UNDP and UFRN:

• The AI ​​solution is able to recommend previous judges in the environmental field, search for similar situations and allow for greater consistency in judgments;

• Processed data containing extracts from environmental causes already processed in Brazil;
• A tool for identifying the largest defendants in environmental causes and pollutants in general from data from the National Judiciary Database (DataJud);
• The AI ​​solution is able to read legal texts and identify important elements, such as the type of crime committed, the damage caused, the biome involved, the value of the conviction and the use of national and international legislation;
• An artificial intelligence solution to predict the outcome of lawsuits in the environmental field.

Text: Raquel Lasalvia
Version: Thais Sieclinski
CNJ . News Agency