Customers: Special equipment and communication of the Ministry of Internal Affairs of Russia NGO (STiS) Product: Artificial intelligence (AI, Artificial intelligence, AI) Second product: Projects - SED - Streaming Recognition Project date: 2021/07
Project's budget: 23 100 000 руб.
|
2021: Tender for a study of the applicability of AI to identify signs of serial seriousness of criminal and administrative crimes
NGO "Special Equipment and Communications" Ministry of Internal Affairs commissioned an applicability study artificial intelligence to identify signs of serial criminal and administrative crimes. Based on the data study, it is planned to create an information system in which technologies speech recognition and handwritten texts, as well as analysis, will be used. large amounts of data This became known on August 25, 2021.
The corresponding tender was placed on August 20, 2021. Applications are accepted until September 13, the results will be announced on September 20. The maximum contract price will be 23.1 million rubles. The study was assigned the cipher "Series."
The purpose of the study, among other things, is to develop a layout and technical task for creating an information system for detecting signs of serial crime. The system will use the machine learning and semantic data analysis technologies selected as a result of the study.
Representatives of the institution "Special Equipment and Communications" of the Ministry of Internal Affairs could not say when it was planned to create the above-mentioned system. They also could not clarify in which other areas the Ministry of Internal Affairs is going to apply artificial intelligence technologies in the near future.
The study under the code "Series" includes, among other things, an analysis of the applicability of speech recognition technologies and optical recognition of printed and handwritten texts, including using neural networks, the tender documentation says.
During the work, the performer will also consider the technology of recognizing entities from text. We are talking about the identification in the text of references to information about individuals (name, date of birth, place of birth, gender, kinship, etc.), their nationality, special signs, as well as legal entities, their addresses and phones. In addition, neural networks will have to distinguish from the text mentions of drugs, dates and times, tools for committing a crime and their characteristics, as well as the method of committing a crime, Internet sites, metro stations, cars, etc.
The object of the study, according to the task, is various sources of information and their digital images. We are talking about the materials of criminal and administrative cases, expert opinions, court decisions, as well as citizens' statements, operational reports and audio materials. The future perpetrator will also have to interview staff directly involved in the investigation of criminal cases. Based on these data, data sets (training materials) for the neural network will be formed.
In May 2021, on the portal of the Ministry of Internal Affairs, it was reported about the agency's plans, using created AI-based programs, to recognize suspicious or stolen vehicles, drivers hiding from the scene of accidents, as well as to search for stolen antiques on open trading platforms and predict criminal events, the[1].