Skip to content

Latest commit

 

History

History
92 lines (58 loc) · 1.93 KB

twitter_recommendation.md

File metadata and controls

92 lines (58 loc) · 1.93 KB

twitter推荐

twitter比较关注social graph的挖掘

1. requirements

use case & product

  • homepage or related item recommendation
  • user: follow
  • item: text, image, video
  • engagement: click, like, comment, share

objective

  • increase the engagement

constraint

  • scale of user and item
  • latency

2. ML task & pipeline

召回、精排、规则多样性重排、混排

  • Fetch the best Tweets from different recommendation sources in a process called candidate sourcing.
  • Rank each Tweet using a machine learning model.
  • Apply heuristics and filters, such as filtering out Tweets from users you’ve blocked, NSFW content, and Tweets you’ve already seen.

3. data collection

  • user
    • demographics
  • item
    • text
  • engagement
    • impression, engagement
  • context
    • device
    • time
  • label

4. feature

  • dense
  • sparse

5. model

5.1 retrieval

  • In-Network召回
  • Out-of-Network 召回

5.2 ranking

  • MaskNet

5.3 reranking

  • 过滤已屏蔽用户的推文、NSFW内容和已看过的推文

6. evaluation

offline

  • recall@k, hit_rate

online

  • ctr

7. deploy & serving

  • batch service or online service
  • A/B testing

8. monitoring & maintenance

9. 优化与问答

reference