1. médialab Sciences Po
  2. News
  3. Socioeconomic dependencies of linguistic patterns and their inference in Twitter

Socioeconomic dependencies of linguistic patterns and their inference in Twitter

Marton Karsai will analyse a joint dataset of Twitter users with an assigned set of socioeconomic features inferred from their geolocation activity. Using this combined dataset he shows how key linguistic variables measured in individual Twitter streams depended on factors like socioeconomic status (SES) , location, time, and the social network of individuals.

Event, Research Seminar

Salle 933, 9 rue de la Chaise, 75007 Paris

Tuesday June 25th, 2019, Marton Karsai will be médialab's guest. He will present his work about "Socioeconomic dependencies of linguistic patterns and their inference in Twitter".

Abstract

Online social networks have become one of the most disruptive communication platforms, as everyday billions of individuals use them to interact with one another. The digital footprints left by users across these media platforms are providing a unique source to study and understand how a user’s language is related to its personal attributes as well as their social network. In this talk Marton Karsai will analyse a joint dataset of Twitter users with an assigned set of socioeconomic features inferred from their geolocation activity: home location was inferred from geotagged tweets and census data was used to assign socioeconomic information for each user. Using this combined dataset he shows how key linguistic variables measured in individual Twitter streams depended on factors like socioeconomic status (SES) , location, time, and the social network of individuals. As an extension he further asks to which degree can these correlated features be predictive of the socioeconomic status of the user and how different SES inference methods might be biasing the prediction outcome.

Biography

Márton Karsai is an assistant professor in computer science at the Ecole Normale Supérieure de Lyon, member of IXXI Complex System Institute. His research interest falls within human dynamics, computational social science, and data-driven research, especially focusing on heterogeneous temporal dynamics, spatial and temporal networks, socioeconomic systems and social contagion phenomena. His main expertise are in analysing large human interaction datasets and in the development of data-driven models of various social phenomena.

Practical Information

Tuesday June 25th, 2019 – 2pm to 4pm.

This seminar is open to all within the limits of available places. Please register in advance.

WARNING, due to renovation works, we had to change the venuethe seminar will take place in Room 933, 9 rue de la Chaise, 75007 Paris