Recommendation System
The system recommends products or contents that may be interested in user behaviour.
The proposal system (recommendation system) is an artificial intelligence system that recommends products, content or services that may attract attention based on user’s past behaviors, preferences or similar users’ images. There are two basic approaches: collaborative filtering (collaborative filtering), "the people who like similar likes but what you haven’t seen yet" suggests, and for this, the user-product uses nucleus in the interaction matrix; content-based filtering suggests similar items to items that the user likes before. These two approaches are often combined with deep learning-based hybrid models.
Recommendation systems are one of the invisible but extremely effective engines of digital economy: suggesting the next series you will watch Netflix, creating Spotify’s playlist, Amazon’s "wholes have also received" recommendations and YouTube’s video stream is always based on this technology; a significant part of these companies’ revenues and user interaction directly depends on the quality of the recommendation systems. On the other hand, these systems are subject to criticisms such as the "filter balloon" effect surrounding the content that only interests the user, so it is one of the important design challenges of the area, to establish balance between discovery and personalization.
Recommendation systems are one of the invisible but extremely effective engines of digital economy: suggesting the next series you will watch Netflix, creating Spotify’s playlist, Amazon’s "wholes have also received" recommendations and YouTube’s video stream is always based on this technology; a significant part of these companies’ revenues and user interaction directly depends on the quality of the recommendation systems. On the other hand, these systems are subject to criticisms such as the "filter balloon" effect surrounding the content that only interests the user, so it is one of the important design challenges of the area, to establish balance between discovery and personalization.