This article is one in a series that explains the main changes we're making to the RIPE NCC website as part of the website redesign project. Here, we take a look at the ways in which we've already improved the search functionality on www.ripe.net, as well as additional improvements that will come into effect when the redesigned site launches in a few weeks' time.
Over the past year, we've undertaken a series of content and development upgrades to improve the search functionality on www.ripe.net . During this time, we implemented four code releases that specifically focused on improving the search experience on our site as part of the larger website redesign project.
This was not a minor undertaking: we have just over 10,000 items within our content management system (CMS), over 40,000 mail items in Mailman, our mailing list manager, and over 5,000 blog posts on labs.ripe.net . All of this content is indexed with our Solr application, and is therefore searchable.
While undertaking this project, the first activity was to analyse our current content and the ways in which people expect to find it. Looking into our web analytics logs has helped us understand what our users are searching for, and the main thing we discovered was that over 90% of the searches on www.ripe.net are searches for an IP address.
During our content review stage, it became obvious that some of our content had been miscatagorised or was missing metadata. Because this data is essential to the Solr application, these documents would often be overlooked in search results.
To address these issues (and many others) we split this part of the project into three sections: cleaning up the content, improving the search algorithm, and improving the search interface.
Old search interface
C leaning up the content
Applied missing metadata to over 900 pages on the website
Conducted internal training to make sure that our content is optimised for searching
Created new internal tools to help our editors maintain the content
Planned regular audits of our content to ensure standards continue to be met
Migrated 250 items in the CMS that were incorrectly categorised with the tag "RIPE" and were being featured in "RIPE*" queries
Added the RIPE Document ID to each RIPE Document, so that searching for a RIPE Document number returns better results
Implemented a content archiving procedure to reduce the results for general searches, during which we archived over 600 old news items as well as 500 meeting minutes, and planned to continually review archiving content when required
Improving the search algorithm
Improved the search functionality when the user searches for a RIPE Policy number
Automatically search the RIPE Database for IP address queries
Improving the search interface
Current search interface
Created a tabbed search view, allowing users to specify whether they are searching for items on the website, in the mailing list archives, or in the RIPE Database
Cleaned up the display to reduce search options that were not being used
Include obsoleted RIPE Documents as default
Clearly display the status of RIPE Documents returned in search results
Quick search option available when a user toggles any of the search boxes
Added the option to search for “archived” content and made it prominent on the search page, allowing users to refine their search when looking for old content that is still on the site but has been excluded from regular search results (such as old news items or meeting minutes)
Inform the user that literal searches often improve results, especially when searching for something with "RIPE" in the title
As part of the website redesign project , we'll be improving search functionality and the interface even more. Here's what you can expect:
Search results that highlighted the search term entered
Ability to search within your current location on the site
As a sneak peek of the new interface, here's what a result will look like for a search of "RIPE 69" on the website:
You'll be able to see the new interface and take advantage of the other coming improvements with the launch of the redesigned website in the coming months.
Feedback
What do you think about the improvements we've already made to the website's search functionality and interface? Please leave your comments here with any feedback you might have.
Mihnea-Costin Grigore is an IT and cybersecurity expert, with extensive experience in areas of software development, networking and project management. He previously held the role of Web Services Team Leader at the RIPE Network Coordination Centre (RIPE NCC), being responsible for the smooth running of the Web Services team, constantly … More
Mihnea-Costin Grigore is an IT and cybersecurity expert, with extensive experience in areas of software development, networking and project management. He previously held the role of Web Services Team Leader at the RIPE Network Coordination Centre (RIPE NCC), being responsible for the smooth running of the Web Services team, constantly … More
Former Web Services Manager at the RIPE NCC, interested in Python, Plone and Pyramids. I've been an open source developer since 2005.
Comments 0
Comments are disabled on articles published more than a year ago. If you'd like to inform us of any issues, please reach out to us via the contact form here.
Comments 0
Comments are disabled on articles published more than a year ago. If you'd like to inform us of any issues, please reach out to us via the contact form here.