w:en:Wikipedia:Wikipedia Signpost/2023-02-04/Tips and tricks

{{Wikipedia:Wikipedia Signpost/Templates/RSS description|1=XTools: Data analytics for your list of created articles: Letting you find out about yourself (and others).}}{{Wikipedia:Signpost/Template:Signpost-header|||}}

{{Wikipedia:Wikipedia Signpost/Templates/Signpost-article-header-v2

|{{{1|XTools: Data analytics for your list of created articles}}}|By PAC2

}}

{{Wikipedia:Wikipedia Signpost/Templates/Signpost-block-start-v2|fullwidth=yes}}

Experienced Wikipedians often have a long list of articles they've created. But what do they know about those articles? How can they get some metrics or analytics to follow their collection?

XTools provides some insights about the list of articles created by a userhttps://xtools.wmflabs.org/pages/en.wikipedia.org/PAC2. The Pageviews API provides a way to get the number of pageviews for each article created by a userhttps://pageviews.wmcloud.org/userviews/?project=en.wikipedia.org. But what about the gender distribution of the biographies I've created? What is the main occupation of people I've written about? Where are the places located for which I've created an article? And if we come to the content of the articles: which is the longest? Which has the most references?

By using the XTools pages-created API, I've developed a set of new tools to answer all those questions.https://observablehq.com/collection/@pac02/pages-created

File:Gender distribution of articles created by PAC2.jpg

At first, I was very curious about the gender distribution of people I've created a biographical article about. So I've used the Wikidata API to get the value of the property {{Wikidata property link|P21}} for all items corresponding to articles a user has created. This first tool is named "User-level gender statistics for Wikipedia".User-level gender statistics for Wikipedia: https://observablehq.com/@pac02/user-level-gender-statistics-for-wikipedia

File:Graph of articles created by PAC2 by instance of.jpg

This tool can easily be extended to other Wikidata properties such as {{Wikidata property link|P31}} and {{Wikidata property link|P17}} (and for humans, {{Wikidata property link|P27}} and {{Wikidata property link|P106}}). This led to another tool named "Look at your list of created articles through Wikidata".Look at your list of created articles through Wikidata: https://observablehq.com/@pac02/look-at-your-list-of-created-articles-through-wikidata

Another tool provides a map of your articles related to geolocated Wikidata items, using property {{Wikidata property link|P625}}.Map your list of created articles through Wikidata: https://observablehq.com/@pac02/map-your-list-of-created-articles-through-wikidata

File:Graph of the number of words and number of references in articles created by PAC2.jpg

We can also gain insights about the content of our articles. The XTools page-prose API gives the number of words, references, unique references and sections in each article. So I've developed a notebook which computes this for all the articles created by a user."Look at your list of created articles with the XTools Page Prose API": https://observablehq.com/@pac02/look-at-your-list-of-created-articles-with-the-xtools-page-ap

File:Insights from PAC2's list of created articles through articleinfo API.jpg

My last tool collects data about the number of revisions, the number of editors, the number of pageviews and the number of watchers for all of a user's articles, using the XTools articleinfo API.Look at your list of created articles with the XTools Page ArticleInfo API: https://observablehq.com/@pac02/look-at-your-list-of-created-articles-with-the-xtools-page-ar

All my tools are developed in JavaScript using Observable (a data visualization platform created by Melody Meckfessel and Mike Bostock), which makes it very easy to design interactive tools. One shortcoming is that you may experience some timeout errors, since my tools rely on a high number of API calls. I can imagine that if you've created more than 2,000 articles, you may have a lot of timeout errors. And all my work is open source – so feel free to improve it and suggest better solutions. And, of course, all of your feedback is greatly appreciated.