Recommendation Engine

A recommendation engine is a type of mechanism that uses machine learning algorithms and suggests the most relevant products to a particular user or client. It works on a principle to find patterns in the behavior data of users that are collected in various ways.

Based on customer's purchase history, most relevant offers are recommended. Also, based of customer's catalog, similar products are recommended. This recommendation can be a comparably higher-end product than the one is in possession by the customer. This type of recommendation is called as upselling.

Similarly, there can be recommendation of products those fulfill additional complementary needs compared to the original items. This type of recommendation is called as cross-selling.

There are two types of recommendation systems:

  • Collaborative filtering: Based on similar customer's past data, recommendations are given to the other customers. Core of collaborative filtering is based on nearest neighbor’s search.

  • Content based filtering: The Content-based approach uses additional information about users or items. This filtering method uses item features to recommend other items similar to what the user likes and also based on their previous actions or explicit feedback.