Research Intern for Discussion Toxicity Prediction (within International Research Project)

We seek reinforcements for our team solving the international project VIGILANT (Vital Intelligence to Investigate Illegal Disinformation), a Horizon Europe project aimed at application of data collection and AI-based web content analysis methods to support police authorities in investigations of disinformation-related crime. In this project, KInIT collaborates with research institutions, software companies, police authorities and training institutions from across Europe.

The intern will work on a specific task of discussion toxicity prediction on social media. The goal is to create a machine-learned model that would predict the scale and intensity of toxicity of discussion that would emerge under a social media post (based on the sole content of that post). In other words, the model would measure how provocative (toxicity indicing) a given post is. The purpose is very practical: when sourcing data for disinformation or hate speech analysis, the social media often limit their APIs to a certain number of requests. Predicting the toxicity would allow to focus the precious API call limits to potentially interesting content.

Your responsibilities will include:
- Creation of AI models and methods, statistical analysis, machine learning
- Data acquisition (including cleaning and preprocessing)
- Data annotation and organization of annotation activities
- Application of machine learned models on practical problem solving
- Experimental software development
- Assisting with execution of research studies

Benefits of the internship:
- Improve your data analysis, machine learning and programming skills.
- Experience best practices of research and project work.
- Opportunity to work on the design of research studies.
- Learn about disinformation ecosystems and methods for their tackling.
- Opportunity to meet the Slovak and European elites in AI and disinformation tackling.
- Work with a very friendly team.

Beyond the VIGILANT project, the toxic content tackling is one of KInIT’s core application domains. Presently, KInIT is a partner in multiple large grant projects, especially in the Horizon Europe program, where we are teamed up with 1st class European research institutions.

Does this internship sound attractive to you?

Read more and apply here with your CV and cover letter.