The other widely used technique is preference-based recommendation. Here, a user is asked to express explicit preferences for certain attributes of the product. If preferences are accurately stated, multi-attribute utility theory (MAUT, [8]) provides methods to find the most preferred product even when the set of alternatives is extremely large and/or volatile, and thus has no problems of sparsity, cold starts, latency or scalability

Leave a Reply

Your email address will not be published.