The recommender system is the fundamental tool in informed consumption, services, and decision-making for effective recommendations. Classic recommendation systems mainly focus on the user-s historical preferences and lack the focus on the user-s latest interaction. A session-based recommendation system is a new criterion of recommender systems that intent to produce more timely and accurate recommendations using short-term but dynamic user preferences for their session contexts. Session-based recommender systems heavily focus on user-s most recent interactions with the items and the user-s incognito session information.
Representation learning is the technique that helps in prediction or decision making, and it allows the system to automatically discover the representations needed for feature detection from raw data. Representation learning in a recommender system supports the system to compute meaningful recommendations in real-time. Incorporating representation learning with session-based recommendation systems determine convenient representation to learn the dynamic information of user session context and result in producing efficient recommendations.