Translations:Course-v1:edz+11+2023 SP/en/block-v1:edz+11+2023 SP+type@html+block@f12e263aa51341c3b407283a280340b9/content/en

There is a lot of data within a Wiki that can help you identify content. However, they each have their strengths and weaknesses. The two main sources of curated connections between article pages on Wikipedias are categories and links. Both of these can help inspire lists of actions or topics, and when combined with Petscan (covered soon), allows you to benefit from the work done on other wikis.

Categories

The most obvious place to look for articles that are similar to each other is categories. Categories are a method of grouping like MediaWiki pages together. Categories act like tags, clustering articles based on a number of characteristics.

<img height="width="100%"" src="/static/image5.png" alt="" />The categories at the bottom of the Climate Change article on English Wikipedia.

However, categories have a number of problems: many newer editors don’t understand how to use them so never update them; the connection between categories and subcategories is often not consistent or intuitive; and they require a large amount of human labor to maintain.

<img height="width="100%"" src="/static/image1.png" alt="" />

The content in the <a href="https://en.wikipedia.org/wiki/Category:Climate_change" target="[object Object]">Category for climate change</a>. Notice how the category has a very odd mix of content -- some of it clearly relevant to the international topic of climate change, but the category also includes generic biographies, country-specific topics, and very niche content. 

Though useful if content has already been well categorized, categories are usually not well used and maintained on different Wikimedia projects. In particular, on smaller language Wikipedias, categories likely will not have a lot of nuance. Also it is very hard to discover whether a category includes all of the content in its scope.

Despite their gaps in information, you can use data from one Wikipedia’s categories to surface content on other Wikis using Petscan. Categories often contain some of the content data on the wikis -- so there is often interesting data, if not always accurate.