🔥Stalecollected in 1m

Bilibili Drops 'Guess You Like' Algo

Bilibili Drops 'Guess You Like' Algo
PostLinkedIn
🔥Read original on 36氪
#algorithm-update#video-platformbilibili-recommendation

💡Bilibili kills core rec algo—signals shift in video ML ranking strategies

⚡ 30-Second TL;DR

What Changed

Guess You Like algorithm fully offline from tomorrow 00:00

Why It Matters

It will launch a new recommendation algorithm instead.

What To Do Next

Review Bilibili's new rec algo docs for ML personalization alternatives.

Who should care:Developers & AI Engineers

🧠 Deep Insight

AI-generated analysis for this event.

🔑 Enhanced Key Takeaways

  • The transition is part of a broader regulatory compliance initiative in China aimed at curbing 'information cocoons' and enhancing algorithmic transparency for major internet platforms.
  • Bilibili's new system reportedly shifts focus from historical user engagement metrics (likes/clicks) to a 'diversity-first' model that prioritizes content discovery and creator ecosystem health.
  • Internal reports suggest the move is intended to address long-standing creator complaints regarding 'traffic stratification,' where the old algorithm favored established influencers over new content creators.
📊 Competitor Analysis▸ Show

| Feature | Bilibili (New Algo) | Douyin (ByteDance) | Kuaishou | | :--- | :--- | :--- | :--- | | Recommendation Focus | Diversity/Discovery | High-Engagement/Viral | Social/Community-based | | Transparency | High (Regulatory driven) | Moderate | Moderate | | Creator Support | Ecosystem-wide | Performance-based | Traffic-sharing |

🔮 Future ImplicationsAI analysis grounded in cited sources

Bilibili will see a short-term decline in average daily time spent per user.
Reducing the hyper-personalization of the feed often leads to lower immediate engagement as users adjust to less familiar content.
The platform will experience an increase in content diversity metrics.
By de-emphasizing historical 'likes,' the algorithm will force the exposure of niche and long-tail content to a broader audience.

Timeline

2022-03
Bilibili introduces 'One-Click for All' and enhanced algorithmic transparency features to comply with CAC regulations.
2023-09
Bilibili updates its recommendation engine to prioritize 'high-quality' content over pure click-through rates.
2025-11
Bilibili initiates internal testing of a non-personalized discovery feed for a subset of its user base.
📰

Weekly AI Recap

Read this week's curated digest of top AI events →

👉Related Updates

AI-curated news aggregator. All content rights belong to original publishers.
Original source: 36氪