Research team will provide guidelines for how to use the topic models and post processing. Updates will be in the parent task.

1. Do you have any known risks in the upcoming quarter?
   • What is the cause? Potential technical limitations, for instance, access aggregated sample dataset in Superset through Presto.
   • What is the impact? If we do not rollout the sample dataset in Superset, we may not be able to get feedback from PMs and iterate datasets in Q3.

Added metrics section for active editors in editors dashboard.

Added diversity metrics for active editors and content interactions in Superset. and add filter for anonymous/non-anonymous edits.

Thank you @JAllemandou !!

@JAllemandou @Milimetric thanks for confirming this with me. I confirm that PMs are aware the data related to platforms are non-additive and they are interested to see editor counts in by platforms. We want to start experiment with creating some test aggregation tables and making sample charts and dashboards in Superset using Presto.

Close the ticket since there's no additional action needed.

@nettrom_WMF can you confirm there's no impact in report/metrics?

Updates will be in parent task.

Maya finished Requirement Document : Content Data Accessibility and shared the requirement document with stakeholders.

@hdothiduc Thanks for the meeting and a great introduction to the project yesterday. And the blog post for the 15th Birthday of Wikipedia is a helpful reference!

In T260369#6434039, @LiAnna wrote:

Thanks! Just to clarify, the usernames I provided were on behalf of the Wikipedia & Education User Group, not Wiki Education, so they include usernames from about 20 global education programs worldwide, not just Wiki Education's Student Program!

Based on the usernames you provided, between July 1, 2019 and June 30, 2020, there are 10.4% (1,887 out of 18,168) of new active editors are from the Wikipedia & Education User Group in an average month.

Red line - internal interactions per external pageview
Purple line - internal pageviews per external pageview

@LiAnna for the lists of username you shared with me, are there any target wikis for these lists?

For both topic dataset, we keep predictions over 0.5 threshold since 0.5 is the official threshold for applying a topic.

Metrics added to content metrics dashboard with filters.

Close this ticket since it's not a bug. We need to document a clearer metrics definitions for Wikistats if we encounter a similar issue again.

Not a bug, the difference is because of metrics definitions. According to this comment, the net new content metric in Wikistats does not include redirects.

Removed COUNT(*) from data sources and data tables in Superset. Didn't see COUNT(*) as metrics in datasets in Turnilo.

Initially, we want to work with Carol as a potential audience and record the training. But the main blockers for her are creating develop account access and availability.
We need to find another candidate who's willing to be recorded.

The drop in new editor retention are from:

• Lower retention rate from European languages wikis, like fr.wiki, de.wiki, pt.wiki and ru.wiki. See this chart in Superset for more details.

Thanks for granting us the access @elukey. I just checked with Maya, both of us can see the user list and update dashboards now.

Updated static axes for charts in editing and readers dashboard.