Recommender system

A referral service (English Recommender System) is an automated process that determined on the basis of existing web pages or other objects similar properties and is recommended. To determine the appropriate recommendations uses a referral service methods of data mining and information retrieval. At the recommendation of hyperlinks also includes information from the concrete context (web access ) and additional information will be used as the purchasing, navigation or click- history, for example.

Types

One can distinguish between four types of context information:

• Current content, ie the currently watched by the user content ( demand paging, product, ... ) and associated additional information such as the category of the content
• Current user, identified by a cookie, and it identifiable information such as Parent information, such as gender, age, origin, ...
• Transaction information, such as previous purchases, previous navigation behavior, areas of interest, preferences, ...

A well-known recommendation engine is that of Amazon.com to recommend book titles and other products. Another example is the service of the KIT BibTip library. In the meantime, take advantage of numerous German companies, the recommendation function, among others, the source GmbH ( prior to their insolvency and cessation of business operation in the year 2009), Metro AG, or movie pilot. The largest academic competition in Recommendation Systems, who was endowed with $ 1 million, the U.S. DVD rental Netflix in 2006 brought into being. The winning solution was delivered in 2009 by a conglomerate called " BellKor 's Pragmatic Chaos". Researchers from commendo research & consulting ( Austria ), AT & T Research (USA), Pragmatic Theory ( Canada) and Yahoo! Research (Israel ) sat there by 50,000 against competing teams.