Big Data Platform: Big Data Music in Russia

Main article: Neural networks (neural networks)

2022: Using a neural network that writes music

On May 24, 2022, Platforma announced the launch of a service to assess the tourist attractiveness of the regions in order to help businesses and civil servants analyze tourist flow and improve tourism infrastructure.

As part of the project's promotion, the developers used a neural network that writes the music of the city using photographs of its horizon line and other geodata, for example, building density and regional attendance. The first five compositions reflect the sound of Moscow, St. Petersburg, Nizhny Novgorod, Arkhangelsk and Kazan. Then the music that the neural network wrote was adapted for listening by artists from these cities - for example, the St. Petersburg Symphony Orchestra - made their own tracks on its basis. The composition is already available on Yandex.Musice.

When creating music for each of the five cities, the neural network relied on several types of data - visual (horizon photo and three-dimensional map of the urban landscape) and population analytics. From the horizon line, the neural network made a waveform so that it was responsible for the timbre. The geodata of the mobile operator Tele2, namely the density of the concentration of residents in the areas of the city, were converted into a sequencer. He put notes in places and in time to make music.

So, the basis for the composition "St. Petersburg" was the line of roofs of Kronstadt and St. Isaac's Cathedral, and in "Moscow" - Red Square and the Kremlin. In the tracks "Kazan," "Nizhny Novgorod" and "Arkhangelsk" managed to use all urban planning ensembles and landscape features.

The second stage of the project was the involvement of professional musicians. Artists of five cities based on music written by the neural network created their own tracks.

The service for assessing tourist attractiveness is based on complex analytics algorithms. The city music project is designed to increase the interest of a large audience in the capabilities of big data analytics and show that neural networks are equally effective in displaying city code in notes and identifying implicit places of attraction for tourists in the region. I want to believe that the service will be in demand from the administrations of the territories, and the created music will create pleasant impressions for residents of cities, "said Anton Merzlyakov, director of big data analytics at Tele2.