Grants:IdeaLab/Namespace monitor/Wikiversité/Département vide
Project idea
editUsing Grants:IdeaLab/Namespace monitor concepts on Wikiversité project, via the Department namespace experiment ; we plan an action to improve this custom namespace.
What is the problem you're trying to solve?
edit- Increase accessibility reducing empty departments number
The namespace experiment below shows the large part of empty departments,see Departments metrics on Wikiversité laboratry.
- Empty department --- 37%
- Department with 3 lessons max. --- 33%
- Department with more than 3 lessons --- 30%
Arguments for empty departments deletion: Impact des départements vides.
What is your solution?
editA process for empty departments, quiet deletion, based on:
- FormWizard to prevent creator & community via talk pages
- Dashboard to monitor deleted departments and associated pages
- Report for discussions and deletions
Goals
editPrior to initial situation based on approximately 2000 lessons:
- 442 departments for 4.5 average of documents per department
See Statistiques sur le nombre de leçons par département.
An optimistic goal is to reach this situation:
- 300 departments near from 6.5 average of doc. per department
- This projection suppose to delete (+/-) 115 departments
The most successfully projection (if everyone approve the process):
- 260 departments near from 7.7 doc. per department
Prioritizing oldest and uncompleted departments ; processing 10 departments per week (quiet an progressive) ; we'll need approximately 6 months to complete the process.
Get Involved
editAbout the idea creator
editPlease, see my local profile on Grants:IdeaLab/Youni Verciti or the global user:Youni Verciti page.
Participants
editEndorsements
editExpand your idea
editProject plan
editActivities
edit- Developing some specific algorithms to return metrics and to provide monitoring for the process.
- Evaluating the situation, elaborating a standard process to perform the task.
- Execute the procedure by:
- advising community and creators on discussion page,
- submit the department and relative pages for deletion (if no negative response from users),
- report the list of deleted pages,
- close the task by publishing the results metrics and reports.
Impact
editMetrics of "namespace monitor" main idea, calculate the average of documents per department, and other indicators (Number of departments... see #What is the problem you're trying to solve? and #Goals). It turns us able to evaluate the impact on the Wikiversity department namespace and documents visibility.
The goal is to reach a maximum proportion of 10% empty departments, it will increase the number of documents per department to 6,5.
I'm publishing namespace's experiences since february 2016 on Wikiversity. Department namespace is the third experiment ; from the beginning the number of departments already decrease, from 449 to 415 ; with help of administrators (as a common user, i just submit pages for deletion). The new process use a FormWizard to advise community on each department's talk page and creator discussion's page.
Example:
- v:fr:Discussion Département:Allergologie Department talk page
- v:fr:Discussion_utilisateur:Wilimut#Proposition de suppressions de département vide User talk page
Or the report for discussions and deletions as seen in #What is your solution?.
Resources
editI am leading with the idea, as an independent contributor. This idea turns-on concrete action the framework "Namespace monitor". I create "département vide" as a sub-idea to provide defined scope and metrics. I request a grant via "département vide" to sustain the widest idea "Namespace monitor", which will keep difficult to explain without example.
- The main part of the resources represent two weeks of development Python & Lua to produce metrics and dashboard.
- List of departments by namespaces Faculté & Département analysing
- Number & average of documents per department
- List of empty departments
- The second part is attributed to design and execution of the process.
- Reporting list of deleted department
- Resuming results, metrics and discussions
Schedule
editThe first works begun early in 2016, the toughest part (coding tools) is done. I estimate that i'll need three months to complete the process by this way:
- Process 10 departments/week
- A month of delay after advising
- Estimation: 120 departments processed
- Beginning: 2016 October 31
- End of processing: 2017 January 31
- Report and close: 2017 February 28
See the french page french page too.