Basic Recommender Systems

Module 1 - Basic Concepts: This module provides an overview of recommender systems, including non-personalized algorithms, global effects, and differences between implicit and explicit ratings. Participants will gain insights into the taxonomy of recommender systems, inferring preferences, and the importance of non-personalized systems.

Module 2 - Evaluation of Recommender Systems: This module focuses on evaluating the quality of recommender systems, covering both online and offline evaluation techniques. Participants will learn about error metrics, classification metrics, and ranking metrics. Ethical concerns and the impact of recommendations on decision-making are also discussed.

Module 3 - Content-Based Filtering: This module explores content-based filtering techniques, including cosine similarity, matrix notation, and TF-IDF. Participants will gain an understanding of improving item-content matrices and the advantages and limitations of a content-based approach.

Module 4 - Collaborative Filtering: This module delves into collaborative filtering, covering user-based and item-based approaches, model-based vs. memory-based systems, and recommendation as association rules. Participants will gain insights into item-based and user-based collaborative filtering and understand the differences between the two approaches.