Krista Lagus

Bio

Krista Lagus received her Doctor of Science (Technology) degree from Aalto University in 2000 in Information and Computer Science, with a minor in Cognitive Science. Her doctoral work was conducted within the WEBSOM research team led by Academician Teuvo Kohonen (PhD thesis, Krista Lagus, 2000) . In this research, she applied neural networks to modeling thematic structures in very large text document collections and contributed to the development of cognitively accessible interfaces for information exploration and retrieval, combining machine learning methods with models of human information processing.

Her research bridges computational modeling with linguistic, cognitive, and societal data. A recurring theme in her work is socio-cognitive modeling: the study of how collective meanings, narratives, and interpretations emerge and stabilize in large-scale communicative environments.

A broader unifying thread across her research is the emergence of representations across levels of analysis. At the cognitive level, she studies how structured internal representations form under informational and computational constraints. At the interactional level, she has examined how shared communicative systems emerge between agents coordinating around a jointly observed reality. At the societal level, she investigates how public discourse constructs shared representations that shape collective self-understanding and social development. Methodologically, her work emphasizes efficient, generalizable, and interpretable modeling of large and heterogeneous datasets.

A related strand of her research applies structured social-psychological frameworks to large-scale textual data. This includes the computational operationalization of wellbeing models and the analysis of fundamental dimensions of social perception and relational orientation. In this work, psychological theories are translated into measurable representational structures in language, enabling the study of how wellbeing, intentionality, and social relations are articulated and negotiated in public discourse.

Her research contributions include the development and application of computational language models (WEBSOM, Morfessor, and others), cognitively motivated modeling of lexical processing, and large-scale analysis of public discourse. She has led interdisciplinary research initiatives connecting wellbeing informatics and structured dialogue (VirtualCoach), and contributed to national research on loneliness within the network led by Professor Juho Saari, participating in the development of nationally implemented loneliness questionnaires in 2011 and 2014.

In the Citizen Mindscapes project, she led efforts to make the Suomi24 dataset available for academic research and examined large-scale emotional and thematic dynamics in social media corpora. An outgrowth of this work, the Medicine Radar project, developed tools for identifying patterns of speech related to medicines and symptoms in social media data. More recently, she has contributed to the continued development of national research infrastructure through FIN-CLARIAH, supporting shared datasets, tools, and open science practices for the digital humanities and social sciences.

Her socio-cognitive research also extends to peace-oriented modeling. In collaboration with Timo Honkela, she contributed to Grounded Intersubjective Concept Analysis (GICA), an approach examining how in-groups develop distinct meaning structures that may underlie conflict while remaining invisible at the level of explicit positions. With a related motivation, Citizen Mindscapes was conceived not only as a data initiative but also as a means of rendering large-scale public discussion analytically visible, broadening recognition of communicative practices often marginalized in formal public discourse.

Research Interests

Socio-Cognitive Modeling of Collective Processes
Interpretable Machine Learning and Natural Language Processing for Social Data
Digital Trace Data and Public Discourse Analysis