Nowadays there is an abundance of information. Last decade there has been an explosion of information and technologies that capture it. Thus, people are facing more and more choices every day. Not only do they vary in the number of possibilities, they also vary widely in quality. The human capabilities of processing information has not seen the same growth of the information available, therefore individuals are unable to rate each alternative and make the best decision for each situation.
To make a choice without any personal knowledge of other options, individuals rely on the experiences and opinions of others. Before a recommendation can be made, an individual should be faced with a decision. Loren Terveen & Will Hill (2001) described this as a choice among a universe of alternatives. As such, the universe is large, which results in the person probably not knowing all alternatives and how to choice between them.
If the individual does not have adequate knowledge to make the decision, he or she may seek recommendation from others. Recommendation are subjective, which means that the recommendation is subject to emotions and opinions. Therefore, recommendations are based on preferences of the recommender. For this research, preference is explained through a decision on a subset of items from the universe of alternatives. Individuals form preferences based on their experience with relevant items, such as a certain height or personality in potential romantic partners.
A recommendation is a resource that helps in making a decision from the universe of alternatives (Terveen & Hill, 2001) and is a way to filter all alternatives into subset of items that is relevant for the individual that makes a decision. We as humans look for recommendations from others who are knowledgeable on the subject and are familiar with the decisions we have to make. One way to do that is by asking close friends or relatives whose views we value. An increasing more popular option would be to use computational recommender systems.
The main goal of a recommender system is to produce suitable recommendations to users of the platform that might interest them. Platforms such as Netflix use these systems to recommend relevant content that is in line with the users’ habits and interests. This results in a reduction of the search effort for users which in turn leads to greater customer loyalty, higher sales and more revenue in general. However, dating platforms differ from traditional content channels that use recommendation system.
Whereas content channels have actual products they can recommend to their customers, the product of dating platform are its customers (and vice versa).