Computational Text Analysis 

Books and server hall.

Computational analysis offers new ways to derive meaning from text. Our project "Mining for Meaning: The Dynamics of Public Discourse on Migration" uses large corpora of text as social sensors to measure what people feel, think, and talk about, which allows us to track the emergence of shared social understandings.

At IAS we have established a research environment committed to the study of the dynamics of public discourse based on large-scale computational text analysis. Generously funded by the Swedish Research Council (grant 2018-05170), the project joins computational text analysis specialists focused on methodological innovation with social scientists attuned to social and political dynamics. This makes IAS a unique place for deriving insights into the evolution of political discourse through the use of theoretically driven computational innovations.

In the wake of digitization, social-media platforms, publicly accessible newspaper repositories, and machine-readable political documents, have become digital archives of public sentiment. Our international and interdisciplinary group---with researchers from Linköping University, Aalto University, the University of Lucerne, and the Institut Politechnique de Paris---analyzes such data to mine for the meaning embedded in these texts.

We study the discourse dynamics between politicians, media, and the public to understand meaning making, explore shifts in sentiment, and follow how collectively agreed narratives arise. In particular, we seek to understand the emergence and transformation of the public debate on immigration in Sweden. Our research will provide answers to questions such as:

  • How do public discourses on immigration form and change over time?
  • What part do social media, traditional media, and political parties play and how do they interact with each other?
  • How does opinion polarization occur online and how does misinformation spread?
  • How do discourse dynamics in local contexts (e.g. neighborhoods) interact with real-world events (e.g. crime) and social outcomes (e.g. residential segregation)?

Our aim is to understand the processes by which issues surrounding migration and integration are framed by different social groups, media outlets, and political parties, which we hope allows for evidence-based policy recommendations that tackle key challenges of integration.

Read more: Who is setting the discourse on immigration

Project time: January 2019 - open end

Our international team of dedicated text analysts come together for a kick-off workshop in March 2019.

Researchers

Activities and output

Dice with letters showing fake / fact.

Sincerely misled: How conservatives and liberals believe online misinformation

We asked conservative and liberal Americans to evaluate true and false messages about politics, society, and science. The results reveal that people often genuinely believe fake news rather than pretending.

Illustration from research data

Seeded topic models for vast text archives

Researchers from sociology and statistics implement a scalable seeded topic model that extracts interpretable meaning structures in perhaps the largest text corpus ever analyzed in the social sciences.

Three glad persons celebrating.

Our first doctoral student completed her PhD

Dr Hurtado Bodell successfully defended her dissertation “Mining for Meaning: Using Computational Text Analysis for Social Inquiry” on 13 May, 2024. This marks the first Ph.D. completion within our research environment.

Funders and partners

Organisation