Security in IoT Social Networks

4. Sources of recommender systems

Based on the source of the data the RS is divided into collaborative, content-, and knowledge-based recommendations.

Collaborative sources: These sources rely on numerous notions to designate the problem field and the specific necessities placed on the system. In the CF technique, the users assist each other to identify fascinating items. Every user rates a subset of items with some scores, so CF does not cover the content description of the items [9]. In this context, the user rating shows his/her willingness or interest toward that item. The CF approach assumes that if the users have agreed with one another previously, they are further probable to agree with one another subsequently over choosing randomly selected ...