Daria Klishevich, an expert from St. Petersburg University, Research Fellow, helped the participants of the Big Challenges research and technological program to develop an analytical service for brokers with recommendations. It connects the change in quotes with world events and allows users to make a considered decision to sell or buy shares. The project was carried out by the team of the Big Data, Artificial Intelligence, Financial Technologies and Machine Learning direction at the Sirius educational center.
The Graduate School of Management of St. Petersburg University took part in the project along with the leading Russian universities — MIPT, Skolkovo Institute of Science and Technology, Novosibirsk National Research State University and others. GSOM SPbU closely cooperates with the Sirius educational center with the support of VTB Bank. We are confident that this collaboration is most useful to each of the parties. Participants of the program have an excellent opportunity to gain modern knowledge from leading experts of the business school and create relevant products under their mentorship, and for GSOM SPbU it is an opportunity to be closer to gifted students and future applicants. After all, together we create knowledge, develop future leaders and change the world for the better”, Olga Dergunova, Deputy President and Chairman of the Board of VTB Bank, Director of GSOM SPbU.
People use different strategies to make investment decisions: they listen to experts, read specialized media and telegram channels. Banks offer "financial advisors" who make recommendations using artificial intelligence. But not everyone trusts such services, because the principle of their operation is hidden from the user.
High school students of Big Challenges have developed their own program that analyzes the news flow, predicts price changes and advises to buy or sell shares. The service also sorts news by topic and shows their tone, allowing users to assess the balance between positive and negative events. The news stream is updated once a day.
“The service has algorithmic transparency for issuing recommendations. A person will be able to read the news and understand what influenced the market, why the stock rose or fell in price. In addition to the buy or sell solution, users get an aggregated summary for any period of time and see a snapshot of the Russian economy: the most important news, key persons and well-known companies. There are no analogs of the service in Russian yet, ” project manager Alexey Ryabykh. Managing Director of Modeling and Data Analysis at VTB.
Participants downloaded data from news sites over the past 10 years — among them were RIA Novosti, Interfax, Kommersant, RBC. Then they analyzed them using artificial intelligence methods: they found proper names in the texts, identified the connections between them, built graphs and classified the texts by topic. They also trained artificial intelligence to make predictions based on the collected data.
“We work closely with machine learning, we create a lot of models and neural networks. I want to work in this area, so I became interested in the technical part of the project. My task was thematic modeling, clustering of texts and their transformation into vectors — the entities with which mathematics works,” Elizaveta Pushkareva, a high school student from Yekaterinburg. She came to the program with a project of a service that diagnoses a stroke using computed tomography images of the brain and tells the doctor about treatment tactics.
Experienced experts helped the team to develop and present the project for the defense. “Under the guidance of colleagues from VTB, students created a news analytics service that helps to understand the relationship between the news flow and stock quotes of Russian companies. The service turned out to be convenient, there is a version for a computer and a smartphone — a serious project in just three weeks. GSOM SPbU specialists together with the students formulated the value proposition of the service, its target audience and what needs it responds to. For us, communication with Sirius's students is always useful, so we can find out what potential applicants are interested in,” Daria Klishevich, Research Fellow of the The Center for Russian Multinationals and Global Business at GSOM SPbU.
Now the team will have to combine the created algorithms on one site, finalize the visualization and improve the performance of the service so that it can withstand a heavy load. At the defense of projects, the schoolchildren presented a ready-made brokerage service.