Beeline, together with Brand Analytics, presented a rating of companies and institutions involved in artificial intelligence in Russia. This was reported by the press service of Beeline. In total, there are 8 main candidates and four special candidates in the rating.
It is stated that the characteristic of the rating is that it is compiled by a trained neural network with data from open sources. Thus, the basis of the datasets were media news and broadcasts of open Telegram channels dedicated to the topic of artificial intelligence for 2022.
Beeline announced that more than 5 billion media materials and Telegram channels were analyzed to search for publications on AI and to prepare datasets. At the same time, out of 1 million publications found, after eliminating duplicates, 500 thousand material strings remained, which were used to train the neural network and create ratings.
According to Konstantin Romanov, Beeline’s director of artificial intelligence and digital products, the company produces and develops solutions based on big data and artificial intelligence on a daily basis. He noted that Beeline regularly reviews cases in the fields of technology development, science, education and industry in order to create customer-oriented solutions that are important on a national scale.
“Beeline’s AI and Digital Products team has sufficient machine learning proficiency to create a system that finds implicit patterns and gives a comprehensive look at the AI market. We wanted to create a rating with Spirit, but remained neutral. “I think we did that,” he said.
As Brand Analytics Managing Partner Natalya Sokolova points out, using machine learning technologies for natural language processing is one of the key areas in the development of the social media analytics industry.
“We are aware that it is important for market participants not only to follow the news in the field of AI, but also to have a traveler in front of their eyes among companies and solutions in this industry, which is important for our future. The project turned out to be interesting. The rating simultaneously takes into account the media activity of the players, the media interest in the subject and the companies, and the “popular” discussion on social media. “Looking at the situation from three angles seems like a very promising approach,” he said.