Knowledge-Based, Ensemble-Based, and Hybrid Recommender Systems

The recommendationRecommendation algorithmsAlgorithm described in the previous chapters used past ratings of users, and preferences of users with similar profiles to arrive at their suggestions of products for new users. But in situations where sufficient past ratings are not available, or there are complex variations in the combinations of preferences of some particular products, the previous methods are not very effective. This is especially true for products that are not bought frequently or high-end luxury products which have high levels of personal customizations or the preferences of users evolve over time. This causes the cold start problem which is a well-known challenge in recommendationRecommendation systems. In such cases to avoid the cold start problem, knowledge-based and hybrid, and ensemble-based techniques are highly useful to give accurateAccurate suggestions.