September 20, 2024

Els Heinsalu and Marco Patriarca,
NICPB - National Institute of Chemical Physics & Biophysics, Tallinn, Estonia

So, every summer various researchers from all over the world come for a shorter or longer period to Aalto University to visit professor Kaski. This means long and inspiring talks at lunch and coffee breaks, interesting seminars during the day, and various brilliant ideas and useful tips that the visitors share with each other. It is a nice way to spend the summer under the pleasently warm nordic sun... and northern lights.

Within the focus of the research activity of Kimmo Kaski's group are some of the most interesting problems on the forefront of complex systems and in particular computational social science and data driven modeling. Among other problems, also the fascinating field of language dynamics that is one of our main research topics, has catched their interest.

The initial research plan concerned the study of geographical names of the Ugro-Fennic area (Els Heinsalu) and the construction and analysis of computational models of personal network formation (Marco Patriarca) and indeed we studied and advanced in both the problems. 

We have analyzed the linguistic correlations between different geographical names. The idea is that correlation between place names may imply some historical correlations between the corresponding places. We have focused first on the data from Estonia. We have found that in Estonia about 1/3 of the geographical names have at least one twin and the same place names are typically not distributed randomly all over the country but within a certain neighborhood. In the future we plan also to analyze separately the toponymic suffixes because they might reveal additional information.

As for the problem of the personal network formation, we advanced in the formulation of the model and its numerical implementation. We are currently exploring how various parameters affect the personal network growth and found that it has a complex dynamics  sensitive to the interplay of different micro- and macro-scopic features such as the type of movements of the individuals, their rates and ranges of interaction, and the size of the domain.

However, not surprisingly, we soon found ourselves discussing about many other interesting problems that led to new investigations. In particular, in collaboration with Tuomas Takko and Kimmo Kaski we started a new data-driven study of the language dynamics in Tallinn. Building on the long experience of the Aalto's group in the computational social science of epidemic spreading, this work aims at realizing simulations of the language contact and competition processes across the districts of Tallinn, taking into account the respective socio-economical and cultural differences as well as the city topological structure and the everyday motion of individuals.

We also advanced on another problem of phone call data analysis started last year, namely the global influence of special events on phone calls, in collaboration with Chandreyee Roy, Kunal Bhattacharya, and Kimmo Kaski. We have proceeded in this problem through a selection and then a focused analysis of some specific events and how they influence the time dependence of the call activity. We have found that an event of general national relevance strongly modulates the pattern and intensity of phone call rates. 

We appreciate very much the time spent at Aalto University. It gave us a possibility to communicate with so many top level scientists and to return home with new ideas and energy that we got not only from the academic environment between the walls designed by Alvar Aalto, but also from the suggestive and calm nordic nature and the Baltic Sea that surrounded us.

Els Heinsalu and Marco Patriarca

Expl

Exploratory: Complex systems, language dynamics