At HSE University–St Petersburg, Experts Discuss How AI Changes Tourist Routes and Cultural Heritage
Neural networks are no longer simple travel assistants; now they decide what to include on the route and what to skip. Artificial intelligence changes how people perceive a place and shapes the image of tourist centres. At the II International Scientific and Practical Conference ’Information Technologies in the Preservation and Promotion of Cultural Heritage,’ researchers discussed who controls algorithmic solutions and in what ways, and whether authentic traditions get lost in them.
The Laboratory for Management in Culture and Tourism has held the conference for the second time within the framework of the university's International Academic Partnership project. Researchers, museum employees, urban specialists, marketing experts, representatives of the tourist industry, and IT specialists gathered at the Rope Shop educational space in St Petersburg. The conference brought together more than 30 speakers from big tech (Yandex.Cloud), museology (Almaty Museum of Arts, Registan), creative industries (Metaforma bureau, Vinci Interactive) and researchers from Russia, Uzbekistan, Indonesia, and China. In four thematic sessions, the researchers discussed the digitalisation of cultural heritage, drivers of territory development, branding, niche tourism, and issues in working with AI.
Larisa Tsvetkova
Deputy Director for Research Activities
Our programme features real-life examples of how technologies help to preserve cultural codes and unite separate digital tools in unified systems. And that's great because the conference is a space not just to share results of your research, but also to launch projects in interdisciplinary teams.
Alena Suvorova, Senior Research Fellow at the laboratory, demonstrated the results of one experiment: researchers asked a foreign model for recommendations of the best places to visit for tourists in Barcelona and then compared them to the recommendations of users on social networks. The researchers put the responses on a map and compared the quantities. It turned out that neural networks offer far fewer places to visit than the recommendation technologies of internet platforms. At the same time, the results of other research show that in some cases, AI assistants focus on non-obvious locations, and in others, they obsess over 'trademarks.'
The researcher also noted that AI recommendations often tritely adjust to the person, taking into account the specified gender, age, and speech pattern. For instance, to women, it often recommends safe routes related to shopping; to the elderly—short and simple walks; and to men whose questions suggest high income—luxury hotels and busy itineraries. As Alena Suvorova mentions, it becomes harder to draw a line and distinguish reasonable personalisation from stereotyping.
Alexandra Pakhomova, Senior Research Fellow of the Laboratory, explores city legends as a form of intangible heritage. In her presentation, she identified a key contradiction: folklore studies strive to preserve oral creative works in their diversity and flexibility. By contrast, tourist marketing refers to simple and memorable images. As a result, orature grows scarce and loses its natural flexibility.
To demonstrate the diversity of city folklore, the researcher presented the project of a database of St Petersburg legends with locations, sources, and classification by plots. Such a body of data, presented as a website, mobile app, or interactive platform, could reconcile the demands of business, technologies, and culture. It records not a unified canonical plot convenient for distribution, but versions available to everyone.
Alexandra Pakhomova
Senior Research Fellow of the Laboratory for Management in Culture and Tourism
If we write down one version of a legend, it is artificial selection. We should record as many versions as possible to present a complicated system of the modern St Petersburg folklore, in which various traditions and phenomena intertwine.
Song Yangyang, Deputy Director of the Institute of Creative Industries Technology at Renmin University of China, is sure that the main danger of working with historical heritage using AI is related to excluding people from decision-making. Without context, computer intelligence doesn't understand what culture means to its members. In this situation, AI, which is prone to average out answers and omit details, can make serious mistakes.
Prof. Song shared his experience of working on a large project: for more than 15 years, his team had been digitising the Kam people's weaving. The experts carried out research in 12 provinces, studying not just patterns and weaving techniques but also people practising the craft: craftspeople, their families, and communes. Based on the collected data, the specialists developed an AI platform which helps artists create ornaments in the traditional graphic language.
Alena Suvorova
Senior Research Fellow of the Laboratory for Management in Culture and Tourism
In the time it takes for research to undergo a full cycle—from research question to publication—neural network models continue to improve, which makes part of the results quickly lose relevance. But general tendencies remain, and the findings serve as the basis for further research, which is something to be glad about.
The Laboratory for Management in Culture and Tourism actively contributes to preserving the cultural heritage of St Petersburg. For instance, this year, under the guidance of Alexandra Pakhomova, students collected more than 400 city legends. Find out more about the projects carried out by the staff on the website.
