TABLE OF CONTENTS

Overview

Collaborative Filtering is a technique used for recommending products that a user might like based on interactions, data, and preferences collected by the system from other similar users.  These algorithms enable you to harness the crowd wisdom to drastically increase cross-sales and overall revenue.


There are two classes of Collaborative Filtering:


User-based: Measures similarity between target users and other users



Item-based: Measure the similarity between the items that target users interact with and other items.



Configuration Examples

Example 1:


Configure this model with the following values:


Type of Collaborative Filtering Model:

Customers who bought this item also bought


1:1 Affinity-based personalization

No


(Optional) Business Rule:


Price > $20


Result: 

Recommend products priced greater than $20,  that were frequently bought by customers who also bought the item.


Example 2:


Configure this model with the following values:


Type of Collaborative Filtering Model:

Customers who viewed this item also bought


1:1 Affinity-based personalization

Yes


Result: 

Recommend products that were frequently bought by customers who also viewed the item. The recommendation results are adapted based on the individual user’s preferences.


Configure Collaborative Filtering


  1. Choose Assets > Strategies.
  2. Click New Strategy.

  3. Choose Contextual  > Collaborative Filtering. The Manage Strategy page is displayed.

  4. Choose the type of collaborative filtering model from the drop-down:

    1. Customers who bought this item also bought

    2. Customers who viewed this item also viewed

    3. Customer who viewed this item also bought

    4. Customers who added this item to cart also added these to the cart

    5. Customers who viewed these items also added these to cart

    6. Customers who added this item to cart also bought

  5. Indicate whether you want to apply 1:1 affinity-based personalization on the model.

Note: If you choose yes, the configured model parameters will be used as a starting point to deliver product recommendations, and thereafter the recommendations will be based on individual user’s affinities.

  1. If you want to optionally include business rules, click Add a new business rule. 

  2. Click Done.