Recommendations in Adobe Learning Manager

Adobe Learning Manager has introduced a new and revamped recommendation system for courses. This recommendations feature uses AI algorithms and users’ interests like Products, Roles, and Levels to provide personalized content recommendations.

The new recommendations system allows you to create custom parameters that learners can select to receive personalized recommendations. These recommendations will be displayed as Courses, Learning Paths, and Certifications to learners on their homepage feed.

To get started with this feature, you must enable the feature in the Admin app.

Enable and configure the recommendations

Course ranking algorithm

The core of the recommendation engine is driven by Learning Manager’s new Course Ranking Algorithm. The algorithm uses 50 million data points and five years of aggregated learning data across millions of users to rank courses based on their likelihood of enrolment. This ranking ensures that most enrollable courses are displayed upfront to the learners.

Key terms

Learning Manager’s new AI-based recommendation engine provides learning leaders with a configurable parameter-based recommendation system for crafting a personalized experience for learners.

The parameters are - Products/topics, Roles, and Levels. Additionally, these parameters can be renamed to suit your needs. So, ‘products’ can become ‘topics’ or ‘roles’ can become ‘region’.

Set up the recommendation system

Adobe Learning Manager’s new recommendation engine simplifies the Admin workflow involved in setting up personalized recommendations because data about Products and Roles associated with a customer/partner is typically available to Admins (for example, from purchase records).

There are primarily three workflows involved in setting up the new recommendation engine:

Admins configure the account’s Products, Roles, and Levels parameter values. For example, an IT solutions provider with banks as their primary customer base may configure the “Product” parameter to have values such as Payment Gateway, Secure Cloud Storage, Fraud Detection System, Trading Platform etc., and the “Role” Parameter to have values such as Integration Specialist, Network Administrator, Risk Analyst, Compliance Officer, etc.

Admins are provided a guided workflow in Learning Manager to optimally set up the recommendation engine and customize the engine based on the account’s use case. Additionally, Admins also get the option of setting up PRL recommendations via a one-time CSV upload.

Data preparation

The user interest data, Product, Roles, and Levels must be uploaded for the recommendations to work properly.

Upload data options

The recommendations feature is configurable. So instead of products/roles/levels, you can use topics/roles/level or choose any of these options,: product/topics only, roles only, product/topics and roles only, roles-levels only or products-levels only.

Based on the recommendation configuration you choose, modify your data sheets accordingly.

The following section explains the most expansive option of using product, roles & levels.

The Administrator must upload user data in a pre-determined format. The uploaded data will then be fed into the Recommendation algorithm, so that a learner receives recommendations for the right courses based on their roles and levels.

Prerequisites

To upload the data for the recommendations to work, populate the Products, Roles, and Levels in the User and RecommendationLO CSVs.

As part of the data preparation exercise, we are providing two CSV templates:

RecUser.csv

The following is an example of records in the csv:

RecLO.csv

The following is an example of records in the csv:

Populate these CSVs and get in touch with your Customer Success team to download the formats & upload these CSVs.

Make the recommendations live

After both the CSVs have been uploaded, click Go live. This will make the new recommendation system visible to learners.

Make the recommendations live

The recommendation system is now available to your learners.

Edit a parameter

Delete a parameter

Delete the parameter

Course settings page

In the settings page of a course, the recommendation for Products and Roles are listed. Learners will be recommended this Course if they have expressed interest in these Products and Roles.

Course settings page

Learner view

For an account with PRL-based recommendations set up, when a learner logs into the learning platform, a guided workflow helps the learner set up recommendations based on his/her product, role, and level preferences. This creates the learner profile for the recommendation engine to analyze.

Learners on accounts that have switched to the new recommendation system can view the recommended courses and training.

The learners can see the following:

After selecting Recommendations on the left panel, the learner sees a popup to set up the recommendations.

Learner sets up the recommendation

Clicking Setup Recommendations takes the learner to the product selection popup.

Select products

Then in the next popup, the learner can select the role.

Select roles

The learner can then add the levels.

Select levels

Learning strips on the learner app

A learner can see the following strips on the app:

Cards on my learning strip

Cards on learning strip

Each card has Rating, Card image, Title, Skill, published date, Author, Duration, progress bar , and Continue or Explore button.

Cards on saved by me strip

Saved cards

Each card has Rating, Card image, Title, Skill, published date, Author, Duration, progress bar , and Start or Explore or Continue or Revisit button.

There is no progress bar on the card after a learner starts the course. A learner can also Unsave the course.

Cards on super relevant strip

Relevant cards

Each card has Rating, Card image, Title, Skill, published date, Author, Duration, progress bar , and Start or Explore or Continue or Revisit button.

There is no progress bar on the card after a learner starts the course.

On the menu, there are two options, Save and Don’t recommend this. If the learner clicks Save, the course is saved to the ‘Saved by me’ strip. If the learner clicks Don’t recommend this, the recommended training is removed from the list.