Portal, V6.1

---

 

The LikeMinds Recommendation Engines

LikeMinds Recommendation Engines communicate with a relational database and generate recommendations. There are three types of recommendation engines:

• Preference engine:

  This engine generates recommendations using patented collaborative filtering algorithms based on users' item ratings.

• Clickstream engine:

  This engine, which also accesses transaction information, generates recommendations based on users' actions as they navigate a Web site; that is, the history of user "clicks" during Web site visits, and the items that users view, click on, and add to their shopping carts.

• Item Affinity engine:

  This engine generates recommendations based on the history of the user's site browsing activity. It matches a currently selected product with a second product that the user would most likely want to purchase along with the first product. For example, if a user is purchasing groceries and adds French onion soup to the shopping cart, the Item Affinity engine could recommend Gruyere cheese to go with it.

• Preference Engine
  

• Clickstream Engine
  

• Item Affinity Engine
  

 

Parent topic

An introduction to LikeMinds

---