education

Education


Round Corner
Department of Computer and Information Science

Oppgaveforslag

Scientific literature recommendation

Many online services provide users with recommendations, to help them find items of interest in the enormous space of available choices. Popular examples of such recommender services include videos (YouTube), music (Spotify), movies (Netflix), online shopping (Amazon), etc. In this project, we wish to provide recommendations for scientific literature to researchers.

ArXivDigest is a platform for personalized scientific literature recommendation. It allows users to sign up and receive a digest email with recommended research articles that are published on arXiv.org. The recommendation part of the platform is open, meaning that external systems can generate and submit recommendations for users, and there is a broker infrastructure in place that combines these.
This project consists of two main parts. It may be pursued by a group of two students or may be split to two individual projects.

  • Part I aims to (i) develop advanced metadata extraction mechanisms (keyword extraction, author name resolution, author affiliation extraction, etc.) and then (ii) use this information to improve recommendations.
  • Part II has the following objectives: (i) employing existing techniques from information retrieval and recommender system for the scientific literature recommendation task, (ii) extending one or multiple of these methods with the ability to explain recommendations, (iii) surface these explanations on the ArXivdigest platform.

Evaluation is to be performed by observing how users interact with the live service.

 

Faglærer

Krisztian Balog Krisztian Balog
Adjunct Professor
248 IT-bygget
 
NTNU logo