Mr. DLib makes it easy to integrate a recommender system into products such as digital libraries, reference management software, or academic social networks. As an operator of such a product, you can use Mr. DLib’s “recommendations-as-a-service”, i.e. your product just needs to request recommendations from Mr. DLib’s restful API and then you display the recommendations on your website or in your application. You do not need any expertise in recommender systems, and integrating Mr. DLib takes between a few hours and a few days, compared to months of development effort when developing your own recommender system.

The typical process of requesting and displaying recommendations with Mr. DLib is as follows:

  1. A partner of Mr. DLib requests a list of related articles (or call for papers, course descriptions, …) for an input document that is currently browsed by a user on the partner’s website.
  2. Mr. DLib generates a list of related articles and returns the article’s metadata in XML.
  3. The partner displays the related articles on its own website.

 

The type of items that are recommended on partners’ websites is chosen by the partners. They may choose to have only their own content being recommended, or content from one of our content-partners such as the Open Access Repository CORE with more than 20 million open-access articles. If you are interested in integrating Mr. DLib into your own product, please contact us.