Researchers Rank Recommendation Algorithms Using Sports Tournament Model
Researchers from the AI and Digital Science Institute at the HSE Faculty of Computer Science have developed an approach for selecting recommendation algorithms more effectively. Their approach uses pairwise comparisons of algorithms to create a tournament table, with the overall ranking based on their performance across all datasets in the tournament. This can reduce the number of algorithms that need to be tested when developing new services, saving both time and money. The study was presented at the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026).
Recommendation systems determine which products, films, songs, or publications to show users. To do this, they analyse users’ past behaviour and predict what they might be interested in next. For example, recommendation systems identify users with similar interests and take into account the sequence of their views or purchases.
However, there is no universal recommendation algorithm. A method that works well for an online store may not be suitable for an online cinema. Therefore, algorithms are usually pre-tested on existing datasets, and their performance is averaged across them. This approach, however, has its limitations: the resulting ranking does not account for the specifics of individual datasets and may be unstable, while testing all algorithms online with real users is costly and risky.
Researchers at HSE University have developed a methodology for comparing recommendation algorithms based on the Bradley–Terry model, which is used to rank competitors based on the outcomes of pairwise ‘matches.’ In this study, the recommendation algorithms were the ‘players,’ while the ‘matches’ were tests of the algorithms on different datasets. Two algorithms were compared using a selected metric, such as recommendation accuracy, and the one with the higher score was declared the winner. Based on the results of all pairwise comparisons, the model estimated the relative strength of each algorithm. The researchers trained and tested 14 algorithms on 89 datasets from different fields.
The results showed that the same algorithms could occupy different positions in the ranking depending on the type of data used in the ‘matches.’ For example, on sequential datasets, where the order of user actions is important, SASRec and GASATF ranked as the top performers. However, when a dataset did not contain an explicit sequence, these algorithms dropped to tenth and eleventh place, respectively, while LightGCN and ALS topped the ranking.
Additionally, the researchers tested the robustness of the rankings to incomplete data. The ranking produced by the Bradley–Terry model remained stable even when some comparisons were missing, meaning that not every algorithm was compared with every other algorithm. The authors also tested an extended version of the model on additional datasets, taking the context into account. In this case, the model had to predict the winners without directly comparing the recommendation algorithms.
'The idea of using a sports model came to us thanks to the work of our senior colleague Vladimir Spokoiny. To draw an analogy with a sports tournament, the outcome of a match is influenced not only by the players themselves but also by conditions such as the city or the weather. In our study, the characteristics of the dataset served as this context, including the number of users and items, the average length of users’ histories, and other parameters. If we train the model to take this context into account alongside the results of previous comparisons, it can use the characteristics of a new dataset to predict in advance which algorithms are likely to perform best,' says co-author Anton Lysenko, Expert at the International Laboratory of Stochastic Algorithms and High-Dimensional Inference.
In 78% of cases, the algorithm ranked first by the model was indeed among the top three. With the conventional approach of averaging performance metrics, this was the case in only 16% of cases. The authors attribute this difference to the fact that, unlike heuristic comparison methods, their model has a sound theoretical foundation, making the resulting rankings more reliable.
'Our approach helps identify which algorithms are best suited to a particular dataset before they are tested online. For example, if a bank needs to recommend loyalty programmes, it can feed the characteristics of a new dataset into the model to identify the most promising algorithms. Only these algorithms would then need to be tested on real users, rather than all available options. This saves both time and resources,' says co-author Sergey Samsonov, Head of the International Laboratory of Stochastic Algorithms and High-Dimensional Inference.
The study was carried out as part of a programme implemented by the HSE AI Research Centre and supported by a grant from the Ministry of Economic Development of the Russian Federation.
See also:
HSE University Wins Autonomous Search Tech Contest for LizaAlert Search and Rescue Team
The navigation system, developed by specialists from the Laboratory for Intelligent Unmanned Systems at HSE MIEM, is designed to search for missing people and allows the unmanned vehicle to autonomously execute route missions without satellite navigation. The Autonomous Search competition is notable for its comprehensive verification of actual working technologies. Out of 50 applications submitted, 15 were admitted to the qualification round, and four teams advanced to the finals.
Road Injuries Leading Cause of Death Among Men Aged 10 to 39
Worldwide, road injuries are the leading cause of death among males aged 10–39. An international team of researchers reached this conclusion as part of the Global Burden of Disease Study 2023, analysing data from 204 countries between 1990 and 2023. Their findings have been published in The Lancet Public Health, with HSE University Professor Vasily Vlassov listed among the collaborators.
Sustainable Management and Business Value: HSE Economists Examine the Relationship in the Russian Market
In the Russian market, sustainable development increases business value primarily when it becomes part of a company’s corporate strategy. This is the conclusion reached by researchers from HSE University and their international colleagues after analysing the management practices of 152 publicly listed Russian companies. The study has been published in the journal Business Strategy and the Environment.
HSE Researchers Find That Doctors’ Ratings Are More Persuasive to Patients Than Experience
HSE University researchers have investigated how doctors’ professional experience and patient ratings affect patients’ trust in medical information. They found that people are more likely to be persuaded by the opinions of highly rated doctors, while paying less attention to their years of experience. The findings have been published in the journal Frontiers in Human Neuroscience.
HSE University to Launch 11 Projects with Research Centres from Five Countries
HSE University has announced the results of its International Academic Cooperation (IAC) competition for joint fundamental research projects. Five projects have received full funding, while six more have been awarded partial funding. The research teams include scholars from Brazil, South Africa, China, India, and Oman. Alexander Balyshev, HSE University Vice President, shared details of the 2026 IAC competition winners.
Leo Tolstoy, Specifications, Counterpoint, and AI Inclusivity: HSE University Discusses Vibe Coding
HSE University hosted the first international conference on vibe coding, VIBE CODING RUSSIA 2026. More than 450 participants and speakers from universities, banks, the telecommunications and cybersecurity sectors, and the vibe-coding community discussed how AI is shortening the path from research findings to products and who is responsible for code written by an algorithm.
Users Lose Out from Restrictions on Personal Data Collection and Processing
An economist at HSE University used a microeconomic model to show that excessively strict restrictions on the collection and processing of personal data can do users more harm than good. The article has been published in Voprosy Ekonomiki.
Window into the Body: Scientists Develop Neural Network to Detect Risk of 15 Diseases from Retinal Images
Russian universities, with the participation of HSE University, Sber, and Z-union, have developed a neural network that can simultaneously assess the risk of 15 types of pathology from retinal photographs, including not only eye diseases but also cardiovascular conditions. The AI system can help clinicians detect potentially concerning changes at an early stage, identify signs reflecting the condition of retinal blood vessels, and determine whether a patient may need further examination. The paper has been published in Frontiers in Medicine.
HSE Researchers Show How Congenital Motor Disorders Affect Brain Development
Researchers from HSE University’s Institute for Cognitive Neuroscience have synthesised the findings of their previous studies on brain development in children with obstetric brachial plexus palsy and arthrogryposis. Their analysis shows that impaired motor function in early childhood not only limits children’s motor experience but also affects memory, categorical thinking, and information processing. The study has been published in Frontiers in Psychology.
AI Users Earn Up to 41.8% More Than Non-Users
Research conducted by economists at HSE University has revealed a significant correlation between the regular use of GenAI in the workplace and higher pay among Russian employees. The study found that individuals who frequently use GenAI in their professional activities earn notably more than those who reject these new tools or resort to them occasionally. The salary premium for highly qualified specialists reaches 41.8%. The article was published in the Voprosy Ekonomiki journal.