Recommender System for Web Articles

Red Hat Developer website brings resources to help developers build modern and innovative apps and services. It includes everything from downloads to developer how-tos and getting-started guides to tutorials, videos, books, and latest new.

Goal of this thesis is the implementation of a recommender system, which brings new, personalized experience into the website. Using machine learning techniques, it should combine user’s website browsing history together with website articles characteristics, and recommend most interesting content to the user. It should provide good balance of the content exploitation vs exploration to the user, which is crucial in the ever changing landscape of the software development techniques and frameworks.

Tasks for the thesis:

  1. Describe the role of recommender systems.
  2. Describe the dataset and compare it with publicly available datasets.
  3. Explore current approaches that are appropriate for building recommender systems.
  4. Choose and describe at least one suitable approach, describe how the selected approach differs from the related work.
  5. Describe your evaluation process for the chosen approach.
  6. Implement the recommender system.
  7. Evaluate the system.
  8. If possible, do an ablation study.
  9. Compare achieved results with state-of-the-art.
This topic is no longer accepting new applications!

Vlastimil Eliáš

Team: Middleware Engineering Services
Location: Brno
Diploma theses with this Topic:
Recommender System for Web Articles