The doctoral dissertation in the field of Computer Science will be examined at the Faculty of Science, Forestry and Technology, Joensuu campus and online.
What is the topic of your doctoral research? Why is it important to study the topic?
My doctoral research develops computational methods for analysing social networks. It focuses on how people connect, interact, form communities, become similar to each other, and build social ties online. The thesis introduces scalable methods for studying large networks extracted from social media. It proposes the Network Strength Index (NSI), a method for comparing tie strength across whole networks rather than individual connections. While I concentrated on the Twitter-based social network, the proposed method is platform-agnostic.
This topic is important because online social networks shape communication, information flow, language change, and social behavior. Traditional social network theories are often based on small, manual studies, whereas today’s digital platforms produce large, complex datasets. My research helps bridge this gap by turning classic social concepts into computational methods that can be applied at scale.
What are the key findings or observations of your doctoral research?
The thesis demonstrates that social behavior on social media can be studied at scale without reducing networks to single connections or simple follower counts. One key finding is that the classic difference between weak-tie and strong-tie networks becomes less visible as networks grow. In language diffusion, once a network reaches roughly 120 nodes, weak and strong networks begin to behave similarly. Another finding is that interaction information is the best signal for measuring perceived similarity between users and outperforms other signals. The research also shows that the Nordic Twittersphere clusters strongly by country, with little evidence of large cross-border or sub-national communities.
The main methodological contribution is the Network Strength Index, a multidimensional measure of the tie strength of entire ego networks. Instead of judging one connection at a time, it combines several interactional and structural indicators to compare complete social networks. The thesis also shows that interaction strength, social similarity, and outlier patterns are the most robust indicators for large-scale tie-strength analysis.
For the scientific community, the work offers scalable tools for studying online social networks, language change, user similarity, and community structure. For the general public, it helps explain how digital connections shape the information we see, the communities we belong to, and the way ideas and language spread online.
How can the results of your doctoral research be utilised in practice?
The results can be used to analyse large online communities more accurately and transparently. The methods developed in the thesis help researchers, public organizations, and platform analysts identify community structures, compare the strength of social networks, and study how information, language, and behavior spread online.
In practice, the Network Strength Index can support large-scale analysis of ego networks without relying on manual coding or simple follower counts. It can be applied to social media research, digital humanities, communication studies, and computational social science.
In addition, the findings on user similarity can improve recommendation systems and help understand why people connect online. Community detection results can also support regional and linguistic studies by showing how online networks reflect national or geographic boundaries.
More broadly, the research provides tools for making sense of digital interaction patterns at scale.
What are the key research methods and materials used in your doctoral research?
The research was carried out as a compilation thesis based on five studies conducted between 2019 and 2025. The main material consisted of large Twitter datasets collected with the now-discontinued Twitter Academic API. These included ego networks, interaction data, tweets, user metadata, a survey-based user-similarity dataset, the Nordic Twittersphere, and multi-region datasets from the Nordic countries, the United Kingdom, the United States, and Australia.
The key methods were computational network modeling, graph clustering, statistical analysis, and similarity measurement. I constructed ego networks, measured interaction patterns, analysed language diffusion, detected communities, and developed multidimensional indicators for tie strength. A central part of the process was designing the Network Strength Index, which combines several structural and interactional measures to compare whole ego networks on a weak-to-strong tie continuum.
Is there something else about your doctoral dissertation you would like to share in the press release?
I would like to highlight the importance of research access to social media data. These datasets are not just commercial assets. They are also records of public communication, culture, language, and social behavior. In that sense, they form part of our digital cultural heritage and should remain accessible for responsible scientific research. However, this kind of access has become much more restricted recently. This creates a serious barrier for future studies in fields such as computational social science, linguistics, communication, and digital humanities.
One striking example is that a dataset collected during my doctoral research would now cost around 3.5 million euros under the current pricing model. This shows how quickly publicly relevant research data can become inaccessible. I believe the scientific community should actively defend fair, ethical, and affordable access to user-generated data for research.
The doctoral dissertation of Masoud Fatemi, MSc, entitled Computational methods for analysing Twitter-based social networks will be examined at the Faculty of Science, Forestry and Technology, Joensuu campus and online. The opponent will be Professor Tuomo Hiippala, University of Helsinki, and the custos will be Professor Pasi Fränti, University of Eastern Finland. Language of the public defence is English.
For further information, please contact:
Masoud Fatemi, [email protected]
