Semantic web is an essential tool for finding and gathering information. However, with so much information available via the semantic web, sometimes it can be hard for users to navigate. Here, I explore the benefits, pitfalls and challenges of semantic web, and explain some of the techniques we’re employing to help users.
I have previously written about how the semantic web can improve user experience and explained that when we create a web of data, rather than of documents, we must think about things more in terms of meaningful real-world concepts, rather than of publishing conventions. So, instead of having to use a traditional volume and issue structure to navigate a site, the user experience can instead be driven by representations of the key concepts that have been semantically tagged in the data. For example, a visitor to an economics website might want to look at debt figures in Estonia and then examine how that varied according to population demographics. The semantic web can make such a topic easier to research. However, it's also worth reflecting on what the pitfalls and challenges are posed by the semantic web for user experience.
Nokia's Ora Lassila hypothesised a while ago as to what some of these challenges might be:
‘After 10+ years of work into various aspects of the Semantic Web and its constituent technologies, I am now fully convinced (read: no longer in denial) that most of the remaining challenges to realize the Semantic Web vision have nothing to do with the underlying technologies involving data, ontologies, reasoning, etc. Instead, it all comes down to user interfaces and usability. Somehow, I repeatedly run into a situation where some use of Semantic Web technologies that would make a nice end-user application is blocked by the fact that the user interface is the real challenge.... For a long time (longer than I have worked on the Semantic Web) I have wanted to build systems that work on users' behalf. Semantic Web is one of the enabling technologies, a means to an end, and not the end itself. Every time I look critically at the current use of (information) technology, I cannot help but wonder how it is possible to actually get away with the approach taken today (where substantial burden is placed on the users).’
This tallies very well with my own experience. The drawback of semantically tagging a large corpus of information is that the results can be overwhelming. To refer back to my earlier example, the total number of all articles on a site relating to debt or a certain geographical territory can often be far more than a user can easily process. Where large numbers of taxonomies are used to semantically tag content, the site structure can become quite fluid and wiki-like, which makes navigation difficult. Similarly, when every single tagged concept is flagged, the results can be quite distracting - creating a large amount of visual noise rather than an efficient reading experience.
So, what options exist to overcome these problems? Here's a few examples of the techniques we're using:
- Card-sorting: a tried and test usability technique for developing groups of navigational items that are meaningful to users. This includes the development of the most sensible set of navigational paths.
I rather suspect that if the UK Government had used this technique when creating their semantic data website, they would have come up with a different set of navigational categories than government department, given that most UK citizens interested in finding out about, for example, government IT procurement spending, would neither know nor care whether this data is published by the Cabinet Office or Department for Business, Innovation and Skills, etc. Certainly the tags users have added themselves reflect a different set of pre-occupations. Which brings me onto:
- Tagging. Given that it's not possible to predict everything that a user might be interested in within a given ontology or taxonomy, allowing users to add tags to content, or suggest where tagging might have been done in error, is a key method of refining and improving content discoverability.
- Faceted browsing and searching. Facets are really great at two things - firstly, recognising that an article or a book chapter is never about just one thing (instead covering lots of different things that can be classified in many different ways) - and secondly, enabling users to take the large lists that semantic web applications can generate and winnow them down to very small lists.
- Concept homepages. Giving each concept in a taxonomy, its own homepage, so there's always a single starting point for a user to look at.
- Highlighting. Allowing a user to highlight key taxonomy concepts within a fulltext article so that they can clearly skim and find what taxonomy terms it's discussing with a minimum of visual noise.
- Auto-suggest. Reinforce filtering techniques like facets by providing a simple auto-generated set of shortcuts to deep site content.
- Showing concepts visually. A lot of concepts that we encounter could often be best shown in visual terms; think of John Snow's cholera map or Henry Mayhew's poverty map and then think how sites like Historypin allow historical data (photos in this case) to be cross-referenced by geography.
Last, but not least, is the point I always come back to; if in doubt, usability testing is the best possible way to find out what your users expect from your site, and how they actually find information.
Look out for Ingenta's new videos about Semantic Web - coming soon to our Blog.
Update: Ingenta will be attending two semantic web events in New York & London in September, organised by Mediabistro. They both promise to provide a great overview of the applications of the semantic web for businesses such as publishers, and we’ve included more information below. Hope to see you there.
Semantic Web Media Summit
The Semantic Web is here and revolutionizing the media industry! Join us at Semantic Web Media Summit, September 14 in New York City and learn how the Semantic Web is changing media production, consumption, and monetization. The event gathers semantic technology and media experts including Mike Dunn (Hearst Interactive Media), Rachel Lovinger (Razorfish), Evan Sandhaus (The New York Times Company), and Mike Petit (OpenAmplify) who will share how the Semantic Web works and what it is doing to transform the media business.
Semantic Tech & Business Conference
Semantic Web Technologies are being used today and creating new opportunities to revamp and build your business. Don’t miss your chance to get ahead of the competition! The Semantic Tech and Business Conference will be held in London on 26-27 September 2011 and will take a look into how companies are successfully integrating semantic technologies and linked open data into their business plans. With two tracks over two days, business and technology experts will explain the inner workings of the Semantic Web and how you can take advantage of it in your enterprise and web-based systems.