Course syllabus

REDI (2017-2B) Research Experiments in Databases and Information Retrieval

Welcome to REDI! Federated social networks may be the answer to scandals with, and abuse of, personal data by large social networks such as Facebook. In this course, students will research how recommendations (such as "Who to follow?" or "Trending topics") can be implemented in federated social networks. A challenge when implementing such recommendations is that in federated social networks, a recommender system cannot use nor know all data in the network.

Students will implement and evaluate a bot on Mastodon that sends its followers (personal) recommendations. Students learn answers to questions like: How do centralized social networks implement recommenders? How to build recommenders with incomplete data? How to evaluate a recommender with user-interaction data? How to write a scientific paper?

Follow #REDI on Mastodon for course updates.

Wishing you a fruitful course,    Djoerd Hiemstra.


Reading Material


Additional reading material


Course summary:

Date Details Due