Cross-wiki Search Result Improvements/explore similar
In the ongoing effort to improve the discoverability of all the world's knowledge, the Discovery team is proposing an enhancement to the search result snippets on the search results page on Wikipedia (Special:Search). The goal of this feature is to surface related content for each individual search result, so that even when the search result itself isn't entirely useful, perhaps its related content could be.
What is it?
- Explore similar would add three new links beside the green metadata on each search result: related pages, categories and languages. When hovered over, these link would expand the search result into a 'card' with related content. These links would reveal the following related content:
- 3 Related Pages
- up to 9 categories
- Any relevant languages the page is available in
- This extended content would be activated by hovering over one of the links. Also, in order to avoid overcrowding the UI, these links would only be visible when hovering over the search result.
How can I see it?
- Logged in users can test this new functionality by editing (or creating) a common.js file - more information can be found in the self-guided testing page.
- It's been said that people only search once, and are unlikely to search again if they don't find what they're looking for. With this in mind, it becomes important to leverage existing search results rather than expecting users to rewrite their query. The feature aims to leverage existing results by offering users related content, so that they don't abandon their search query entirely.
Can I help?
- Yes, please!
- Update your common.js file in your logged-in profile and try out the functionality and let us know what you think of the results. Are they interesting and/or useful; do they help with your research or editing needs; do they pique your curiosity about information related to your original search query? Feel free to add questions and constructive comments in the talk page.
- We've identified a few communities that might be interested in providing early feedback for this new functionality. We plan to reach out to these groups to get feedback before further deployment.
Hasn't something like this been done before?
- Yes, there is a similar extension that was created as part of a new project for related pages. This feature was first introduced in the mobile apps and as a beta feature on desktop and mobile web.
- Initial community feedback was to: not appear for disambiguation pages and only use non-free images in certain circumstances. This is the current phabricator ticket for related pages on mobile.
Explore similar - for languages in search results
In the ongoing effort to improve the discoverability of all the world's knowledge, the Discovery team is proposing an enhancement to the search result snippets on the search results page on Wikipedia (Special:Search). The goal of this feature is to highlight that the articles returned in the search results can be read in other languages. This will be displayed with an icon and the text 'Languages'.
What is it?
- This is a secondary and modified A/B test from the first one (noted above). The languages displayed will be dependant on the user's browser preferred languages and then the languages that are common in the region that the user is in (using the Universal Language Selector tool). Not all search results will have additional languages that the article can be read in.
How can I see it?
- Logged in users can test this new functionality by editing (or creating) a common.js file. Not sure if you have a common.js file? Make sure you're logged in and then use this url and see if the common.js file needs to be created:
- Go ahead and create it if you need to and then add in the following lines and save your common.js file.
importStylesheet( 'User:JDrewniak_(WMF)/exploreSimilarSearchResults.css' );
importScript( 'User:JDrewniak_(WMF)/exploreSimilarSearchResults.js' );
- Now, when you get the search results page from a query, you'll see the explore similar language link with each search result.