Meta explains in detail how AI selects content recommended on Facebook and Instagram
Meta uses AI to recommend content that matches users' tastes on Facebook and Instagram, which are SNSs operated by the company. Meta has released detailed information on how AI recommends content on these SNS.
How AI Influences What You See on Facebook and Instagram | Meta
The AI behind unconnected content recommendations on Facebook and Instagram
https://ai.facebook.com/blog/ai-unconnected-content-recommendations-facebook-instagram/
Meta explains how AI influences what we see on Facebook and Instagram - The Verge
https://www.theverge.com/2023/6/29/23778068/meta-facebook-instagram-social-media-algorithms-ai-transparency
Nick Clegg, president of international affairs at Meta, said: 'With powerful technologies like generative AI advancing so quickly, it's understandable that people would be both excited about the possibilities and worried about the risks. I think the best way to address these concerns is to be open,” and released information about AI being used on Facebook and Instagram.
As part of this, a page called 'Approaches for explaining rankings' has been published on Meta's Transparency Center . Here is a detailed and easy-to-understand description of how AI is used in the core functions of Facebook and Instagram.
Our approach to explaining ranking
https://transparency.fb.com/features/explaining-ranking
For example, in the 'Explore' tab, which is 'the ability to view photos and videos of accounts you don't follow' on Instagram, the three steps taken by the automated AI recommendation system are as follows: It is explained.
1: Content collection
The Discover tab's AI recommendation system collects content (photos, reels, etc.) published on Instagram that adheres to Meta's quality and integrity rules .
2: Utilization of signals
AI recommendation systems take into account 'how users engaged with similar content and interests'. This is called the input signal.
3: Rank your content
The system ranks the content from Step 2 above and pushes content that it predicts will be of most interest to users to the top of the Discover tab.
According to this page, Instagram users can change the trend of recommended content displayed in the Discover tab by taking either 'Save' or 'Not Interested' actions on content displayed in the Discover tab. It seems that you can do it.
In addition, Facebook's '
Meta also announced that it will begin rolling out a new suite of tools for researchers, Content Libraries and APIs, in the coming weeks. This content library and API contains tons of Instagram and Facebook public data that can be searched, discovered and filtered. Meta explains that these tools can be applied for access through approved partners, including the Interuniversity Consortium for Political and Social Studies at the University of Michigan.
Meta explains that these tools are 'helpful in meeting compliance obligations regarding data sharing and transparency.'
Related Posts:
in Software, Web Service, Posted by logu_ii