Recommended Content
Percipio uses artificial intelligence (AI) to provide recommendations. You see recommendations on the home page and in email notifications. There are three recommendation strips on the home page, each can contain multiple tabs:
- Top picks for your role: If you selected a role in Percipio, you see a strip labeled with that role. View recommended content by content type for skills mapped to your selected role. Every week, we select 3 skills that are associated with the role and display the top 5 search results based on those skills. You can find the skills that serve as the basis for the recommendations by selecting the information icon. The skills change weekly ensuring that content recommendations stay fresh and relevant to help learners continually grow their career!
- Based on your selected skills: For each skill that you select, Percipio shows a tab that contains recommended content. Percipio presents new recommendations weekly based on the most popular (Liked) and newest content from the selected skill interests.
- Recommendations for you: This strip contains up to five tabs depending on what content you have entitled. You may see:
- Recent activity: If you recently started any content, Percipio shows this tab. Skillsoft uses a technique called Collaborative Filtering to recommend items on this tab for you. This technique uses machine learning to analyze anonymized activity data from all Percipio users and create an AI model. This model is used to predict what content you might be interested in, based on the content that other users who accessed the same content also accessed. This is similar to the ubiquitous 'People who bought this item also bought...' type of recommendation.
The Collaborative Filtering AI model relies on content usage patterns of other users, so there are some cases where it does not work well, for example, newly released content, which doesn't have much usage data, or custom content, which is only present in a single Percipio org. In these scenarios, Skillsoft uses Google BERT vectors to recommend content based on contextual similarity. This means Percipio will recommend content that is most similar to content you just accessed.
If you are not interested in a particular item on the Recent Activity recommendations, you can select Not Interested from under the action menu, which removes the item from the strip.
- Trending: Shows the most popular content in your organization based on unique user activity from the prior week.
- Business Top-Rated: Shows content that has received the highest rating from other users if you are entitled to the Business collection. Recommendations are only available for English courses.
- Technology Top-Rated: Shows content that has received the highest rating from other users if you are entitled to the Technology and Developer collection. Recommendations are only available for English courses.
- Leadership Top-Rated: Shows content that has received the highest rating from other users if you are entitled to the Leadership collection. Recommendations are only available for English courses.
- Popular Channels: Promotes entitled channels based on aggregated Percipio activity. Popular channels display in the selected Percipio site language.
- Recent activity: If you recently started any content, Percipio shows this tab. Skillsoft uses a technique called Collaborative Filtering to recommend items on this tab for you. This technique uses machine learning to analyze anonymized activity data from all Percipio users and create an AI model. This model is used to predict what content you might be interested in, based on the content that other users who accessed the same content also accessed. This is similar to the ubiquitous 'People who bought this item also bought...' type of recommendation.