The project is no longer available

User needs modeling in the newsroom

Users engage with news in different ways based on their needs, e.g. need for updates, perspective, inspiration, entertainment etc. Increasingly, media outlets attempt to cater to different user needs in news personalization. NLP has the potential to assist in this process by classifying news articles wrt user needs.

Bildet kan inneholde: diagram, symbol, grafikk.

Members of LTG are partners in the center for research-driven innovation, MediaFutures (hosted by University of Bergen), alongside a number of Norwegian media outlets (Schibsted, Amedia, DN Group and TV2). This thesis will work with ML developers at TV2 on modeling and evaluation of NLP-based user needs modeling. Depending on the interest of the student, the thesis could address one or more of the following topics:

  • (multi-label) classification of articles based on user needs
  • LLM-based fewshot classification
  • model distillation for efficient user needs modeling
  • evaluation/benchmark creation
  • assessing data variation, in particular classification of transcripts from the newsroom

The thesis supervision will involve Lubos Steskal from TV2 on the supervisory team and the student will have the possibility of having one or more internships at TV2 during the thesis work.

 

 

Published Oct. 1, 2025 7:16 PM - Last modified Oct. 5, 2026 4:00 PM

Supervisor(s)

Scope (credits)

60