Russian Students Build AI Navigator to Rank Service Quality
An AI-powered system analyzes user reviews to recommend the best-rated venues—without bias.
A team of Russian university students has built an intelligent service that helps users choose the best-rated venues and services based on authentic customer feedback. The innovation was developed at the Moscow Aviation Institute (MAI). According to Ilya Vedmakov, a senior lecturer in the Department of Computer Systems and Networks and a lead engineer on the project, a single search query can now reveal the top-rated tire shop, hair salon, café, or any other local service. The AI model sifts through user reviews and ratings to provide data-driven recommendations, stripping out subjective bias.
Smarter Review Analysis
The developers found that people tend to overrate experiences, usually clustering between 3 and 5 stars. This distorts the overall picture of service quality. To fix that, MAI researchers introduced a three-tier classification system. The AI now analyzes each review and labels it as negative, neutral, or positive—offering a far more accurate reflection of customer satisfaction levels.
Only Authentic Feedback
Currently, the algorithms are trained on anonymized datasets from Yandex.Maps. In the near future, the system will integrate data from other popular platforms, creating a broader and more representative dataset.
Soon, the AI will be able to assess not only sentiment but also the core meaning of each review, boosting the precision of its recommendations. Researchers are also developing methods to detect fake reviews—an essential step in combating businesses that inflate ratings for reputation gains. With this innovation, such manipulations could soon become a thing of the past. A recent Yandex.Market study revealed that 75 percent of Russian consumers trust online reviews when choosing products.
