finding items that have similar purchase profiles, or finding users with similar profiles. When a new user enters the picture, or when a new item enters the catalog, the recommendation algorithm has nothing to work with. In such a case, various acceleration strategies can be used, most relying either on recommending something that is generally popular (though this makes it difficult for new items to garner recommendations), or using item attributes to recommend something similar to the purchased or rated items.
Cold Start Problem
There are various ways to recommend additional products to an online purchaser, and the most effective ones rely on prior purchase or rating history -