Facebook Suggested Groups AI system

UPDATED MAR 7, 2025
The content you see on your Facebook Suggested Groups is selected, ranked and delivered to you by an artificial intelligence (AI) system. Within one AI system, multiple machine learning models work together to deliver your experience. These models and their input signals are dynamic and they change frequently as the system learns and improves over time.

Overview of Facebook Suggested Groups

When you view and interact with Facebook, one of the underlying AI systems suggests public or private groups to join that you may be interested in. These suggestions primarily show up when scrolling through your Feed.
How Facebook Suggested Groups works
The AI system behind Facebook Suggested Groups automatically recommends groups you might find interesting and useful. Here’s how it works:
  1. Gather inventory
    First, the system gathers relevant public and private groups that might interest you. These might include groups your friends have joined or groups that are related to topics or products you’ve recently engaged with. We have many generators to retrieve group candidate seeds.
  2. Apply additional rules
    Then, the system applies integrity processes to every group to ensure that only groups that meet our Community Standards are recommended. At this stage, the system also applies location and language filters to ensure that relevant groups appear. Then, specific rules filter out groups that have a low number of members.
  3. Score groups
    Finally, the system calculates a score for each group based on a variety of factors, such as how well the group matches what you tend to interact with on Facebook, how often people join the group or how many people are in the group.
How to customize what you see
Your experience on Facebook Suggested Groups is personalized based on your activity, and you have options to control or customize what you see. Below, we describe how to do this with different in-product features. Options shown here may not be available to everyone.
Discover non-personalized groups
To discover groups and posts from public groups that are not personalized specifically to you, use the search feature in the Feeds tab. By doing so, your search results will be based on the search term you enter. Learn more.
Hide suggested group
You can hide a suggested group you’re not interested in. This helps to ensure that similar groups won’t be recommended to you.
How the AI delivers content to you
We want you to see content you enjoy and find interesting. To achieve this, the AI system has models that help it make predictions about content you'll find most relevant and valuable. These prediction models use underlying input signals to help select content you're most likely to engage with.
Below are some of the significant predictions–and input signals that inform them–that we use in this AI system.
How likely you are to view News feed post
Signals influencing this prediction include:
  • Your News feed post viewport views
How likely you are to cross out a group
Signals influencing this prediction include:
  • Groups you have crossed out
How likely you are to join a group
Signals influencing this prediction include:
  • How many members in total across all groups you are a member of
  • How many groups you have joined
  • How many active groups you are a member of in the past day
  • The position of the group in your set of recommended groups
Signals influencing this prediction include:
  • Similarity between user and group members
How likely you are to skip a group
Signals influencing this prediction include:
  • How many groups you have skipped
How likely you are to view the post
Signals influencing this prediction include:
  • Your linear viewport views
How likely you are to engage with groups
Signals influencing this prediction include:
  • Your group engagement metrics count
  • Your HTE value and count
  • Your UHTE value
How likely you are to match the group intent
Signals influencing this prediction include:
  • Group intent
How likely you are to comment in the group
Signals influencing this prediction include:
  • Your comment count