Fluidic Navigation

I have a dream – to think as little as possible. I agree with Arthur C. Clarke when he suggests that “Any sufficiently advanced technology is indistinguishable from magic.”

Therefore, how can we get technology to “magically” make things easier for us to do. The invisible interface, which I have covered before, is one outcome of such a concept.

In this particular case, however, I am referring to web navigation. In some ways, arguably the biggest user-interaction challenge over the medium of the web.


There are several parts to this challenge:

  • How can you make sure that your visitors can get to the information they are looking for in the shortest amount of time possible?
  • How can you ensure that other information that is relevant to them is found (or stumbled across)

Of course, if the amount of information being provided is small, the above concerns do not apply – the visitor could go through the whole website in a matter of minutes as long as there is clear navigation present.

This becomes more and more critical the larger the amount of information

From this perspective, instead of having fixed navigation, since every visitor is different, how would one build a fluidic navigation based on the areas in which the user has expressed an interest in.

Building it

Assuming that it is possible to build a fluidic navigation,

  • How do you capture the users areas of interest without asking them outright?
  • How do you them find content that is similar and could be relevant?
  • How do you provide the user with the ability to navigate to these pages?
  • How do we track whether the navigation options that were given are, in fact relevant?
  • How do you measure the effectiveness of the system?

Even with just the basic concerns, it quickly becomes a complex system to create with challenging problems to solve.


Ensuring the privacy of the visitor is another key aspect. The system has to collect and store (potentially indefinitely), a lot of information on how the site is used. This will primarily include metrics of navigation patterns, not far off from information usually already collected by web server logs and tools like google analytics. In most cases, the navigation patters are not associated with an individual person.

For websites that have user registration and logging in options, there is possibility of associating this information with an individual.

There is however, the question of whether and how much of an invasion of privacy, the collation of this information is. In any case, there is also the question of how relevant it is if the information is only accessible and relevant to automated systems.

Semantic Web

The semantic web (or Web 2.0), was brought into play to resolve a subset of challenges that this kind of navigation brings. For example, when a visitor is viewing a page about catfood, how relevant to them is Cats, the musical?

A Real World Problem

kraya, as a organisation has grown from a two-person home based company into something that has five specialist teams each concentrating in a specific area of technology and/or creative sector.

We have several target markets, existing clients, potential clients, potential team members and members of the community.

As big proponents of innovation, the open source community and the creative sector, it is very important for us to engage with people and organisations in a variety of sectors, geographic locations and languages. There is also the requirement that there are different mechanisms of communication and collaboration that will be used. Of course, the site is also expected to be used from a range of clients and by people who have impairments. As an exceptionally diverse group of people working at kraya, it is not enough for us to just to include them, the site should work just as well (relatively) for everyone.

With such a diverse set of information to be disseminated to a range of people and organisations, from a whole range of different perspective, it quickly becomes a complex and mammoth task.

There is also a lot of conflicting requirements. For example, krayamotion specialises in audio-visual media. The only way to provide access to this information is through flash or embedded audio/video files. How do we then make sure that people without broadband are not excluded? What about people with visual or auditory impairment? If we build the site for maximum accessibility, how do we then ensure that the quality of the work is effectively communicated – how can we still convey the wow factor?

As a small organisation, exceptionally busy with service delivery, this challenge was insurmountable.

Solving it

To solve this problem, we only had one choice. Start small. In fact, we started tiny. To address all of the requirements, it would take a team of 10 or 15 highly capable people roughly 18 months.

We had access to roughly three highly capable people for about 3 person months spread across 5 months.

Why reinvent the wheel

In my view, there are already three different websites out there that address this in three different ways. Amazon builds together recommendations from previously purchased items and your ratings, along with similar items based on items you are viewing / have viewed.

Wikipedia does a good job building navigation based largely on in-content links and navigation.

Google provides search that excels at helping you narrow down relevant information easily.

With these in mind, we decided to pull together a system that uses all three.

What we used

For the web platform, we are using drupal, along with a whole range of plugins including taxonomy and domain access. Some of the plugins where modified, sometimes heavily to provide a specific set of functionality.

Tagging pages is another challenge since it is still a manual process. The is integrated with open calais, also used by tagaroo for WordPress.

Where we are

Six related websites are now live platform for the addition of more functionality. The kraya homepage gives guidance to the different websites. While there is a dropdown bar for traditional navigation, it is expected not to be heavily used.

What the future holds

We are looking to integrate learning algorithms, probably using something like neural networks. The system will also require fairly extensive information collection capabilities to measure how well it works and obtain feedback on how the mechanisms work. Step by step, over the next 12 to 18 months, the site will get smarter and easier to use.

In the meantime, if you have any comments or feedback, please do not hesitate to get in touch, either via this website, or the contact forms on our website.

WordPress Themes