Loading...

Solving the Cold-Start Problem in Recommender Systems Personalization

Maheri, Mohammad Mahdi | 2023

104 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 55864 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Rabie, Hamid Reza
  7. Abstract:
  8. User cold-start is a common problem among real-world applications in the sequential recommendation field since determining user preference based on a few interactions is difficult. The problem would end up limiting the performance recommender systems. To address the cold-start problem, some previous works used meta-learning along with user’s and item’s side information. Meta-learning algorithms made the model able to share knowledge among all tasks. Although they had promising results, they had some fundamental issues with modeling the dynamics of user preferences and considering all kinds of users’ preferences, especially for minor users. The proposed method includes a model incorporating dynamic information in the sequence of users to predict the following items more accurately without needing side information. Also, based on the proposed model, the fundamental and challenging problem of using meta-learning in the recommendation problem is addressed. Through that, minor users’ preferences would be considered without being collapsed by major users. In addition, users will be clustered to take advantage of other users’ knowledge that exists in the same cluster. In different experimental settings, the proposed method has been validated against multiple benchmark datasets. Based on empirical results, the proposed method consistently outperforms several meta-learning recommenders. By doing that, the proposed method improved 4 to 16 percentage improvement on MRR metric compared to state-the-art works
  9. Keywords:
  10. Personalization ; Suggestion System ; Recommender System ; Cold-Start Problem ; Metalearning

 Digital Object List