IPA: Challenging the Traditional

When your website contains thousands of pieces of content, it is sometimes a good idea to step back and re-evaluate how effective your navigation is and how easy it is for your visitors to find the content you’ve provided for them.

One of the overriding reasons the IPA had for redeveloping their website was the challenge of presenting a vast library of content, made up of thousands of news articles, documents and articles. This has always been a major issue for the IPA’s where attempts to deliver this library through traditional navigation have been largely unsuccessful.

We rose to that challenge with our traditional enthusiasm at the same time issuing our own challenge by asking the IPA to take a brave step away from traditional navigation; we suggested employing a number of methods usually associated with social networking and image sharing sites to enable site users to discover and explore relevant content.

At the back-end, the submission and publication process revolves around the key wording and categorisation of all content; it is also possible to link related content together. With content set up in this way, we were able to offer a series of exciting and useful tools to the IPA, each on their own exciting; combined they would offer an extremely powerful and flexible user experience.

Search

The first thing that required tackling was a comprehensive search facility. We designed our own bespoke weighting system that would take a wide range of criteria into account when determining how relevant a piece of content was to the search term used. These criteria include word order relevance, date and location of search term (title and keywords being highest).

Tag Cloud

Simply providing a search engine was not enough; we began designing and developing a keyword thesaurus within the system allowing the IPA to set up synonyms (relationships between words and phrases) that can then be applied to content.

The front-end of the website would reflect these choices threw a ‘tag cloud’. The tag cloud works in several ways:

• Firstly showing the actual keywords associated with the content;
• Then identifying what additional synonyms had been associated with those keywords;
• Finally calculating the relative frequency of each keyword (or tag) within all content
  

The more frequently the tag appears in content, the more prominent that tag appears in the cloud. Clicking each tag in the cloud returns a new set of matching content and the tag cloud redraws itself to show yet more relevant tags.

Categorised and Related Content

As part of the process of creating or editing content, the IPA can ‘group’ articles of the same subject, providing another very simple solution to the issue of delivering large amounts of content. Whenever content is associated with a subject, the website shows a link in its navigation that lists all content of the same subject.

Aside from this wealth of dynamically generated associations between content, the IPA can also link specific articles within the site. Once a relationship is formed between two pieces of content, the CMS automatically ‘reciprocates’ that relationship, placing a link on both.

The system also dynamically removes links if content is deleted, suspended, expired or otherwise removed from the website.

The IPA website is currently undergoing usability testing; you will be able to witness these extremely clever solutions first hand when the IPA website launches to the public in early 2008.