Description

Within the Attribution API, we are surfacing two different types of trending:

• Top: Reflects that the page is one of the top N pages read or edited articles across the entire wiki project. An example might be a world event like a celebrity death, or a persistently popular topic like information about global leaders.
• Relative: Reflects that the page is performing more strongly than usual, but may not be a top read page overall. An example might be a newly breaking news story or relatively niche topic showing stronger than usual interest, such as a new season of a lesser-known anime being announced.

Relative trending requires a nuanced model to detect, in closer-to-real-time, whether a page is trending. However, top is already pre-calculated on a daily basis through the Analytics API. We should utilize the pre-calculated values to start, while conversations around trending more holistically evolve.

Conditions of acceptance

• Implement the approach suggested through the research task for this request: T418501: [SPIKE?] Attribution API MVP: Determine how to add 'top' trending signal to the attribution trust_and_relevance object --> This ticket may need to be broken down further to establish the maintenance script separately from the real time lookup
• Update the top.edited description to indicate that it is updated monthly:
  • Suggested copy: Indicates the page is currently in the top 100 edited pages for the wiki project, based on number of edits. This value is updated every 30 days.
• Update the `top.read' description to indicate it is updated every 24 hours:
  • Suggested copy: Indicates the page is currently in the top 1000 read pages for the wiki project. This value is updated every 24 hours.
• Update the 'top' description to remove 24 hour emphasis
  • Suggested copy: Indicates whether a Wikipedia article ranked among the project’s most read, most edited, or both over the last 24 hours
• Notify Halley & the future audiences team when the work is complete