IPGems Logo Presentations list
s_line.gif (886 bytes)
IPGems home page
s_line.gif (137 bytes)
Resources on Usability and Performance-Centered Design
s_line.gif (137 bytes)
Resources on Information and Knowledge Management
s_line.gif (137 bytes)
Contact IPGems
s_line.gif (137 bytes)

User Interaction with the Semantic Web

It has become more practical, and also more beneficial, to build semantically-enabled applications.

The Semantic Web offers opportunities to achieve designs that require strong context (i.e. knowledge of the user and situation), integration of information, and most of all a key focus on the user. This is what first drew our interest in 2001. Since 2004, we have been collecting examples of current innovations related to user interaction/design for the Semantic Web.

This page serves as a reference for usability/design practitioners, developers, and project leaders as you explore the subject and consider the role of Semantic Web opportunities and offerings in your work. To help get oriented, you can view the latest presentation slides (June 2008) from the Usability Professionals' Association conference, and others from Computer Human Interaction (ACM SIGCHI), May 2007 (pdf). Examples and interesting web sites are listed further down the page.

At the same time as the Semantic Web becomes more practical for interaction designers, academics and practitioners from the Semantic Web community are seeking the involvement of usability professionals to continue defining the "next generation web."  Workshops and papers at most conferences have increasingly focused on user needs and user interaction.

This page includes:

Background

The Semantic Web has been defined as "an extension of the current web in which information is given well-defined meaning, better enabling computers and people to work in cooperation" (Tim Berners-Lee, James Hendler, Ora Lassila in Scientific American, 2001).

The tools for the "next generation web" are becoming available now and will be increasingly used by web developers, so it is important for user-centered designers and usability professionals to understand what is coming into greater use.

If we think of the challenges and limitations we have when designing currently to meet user needs, then it appears that the Semantic Web is dangling some enticing carrots in front of us:

  • Greater user context available to make application/site behavior more user-centered
  • More flexible and adaptive language parsing and "understanding" to make the conversation with the computer more effective
  • Better visualizations and representations of complex data and interactions
  • Availability of agents that can manage routine background tasks and reduce the need for users to hunt and search among thousands, millions, or billions of items of information without adequate support
  • Sharable ontologies and vocabularies, allowing all the applications and sites that a user sees to speak a "similar language" and reducing the demand on the user to interpret language when trying to complete a task that requires more than one site or application at a time
  • Logical inferencing that can help synthesize information from disparate sources, compare and organize that information, then present it to the user in a more seamless way, reducing the cognitive burden

While it entices us, it also challenges us to consider where and how "semantic enabling" user interactions can facilitate increased usability of the web and software applications. It challenges us to consider the risks and complications that arise from these technologies, and the role we need to play as user advocates and representatives in the ongoing development of these technologies. Some of these challenges are explored in further detail in our 2004 paper: the Usability Imperative Inherent in the Semantic Web.

return to top of page

Where to focus? Interaction challenges for the Semantic Web

Below are some broad categories of challenges that need both research and practice input from the usability community. A growing list of challenges is being developed by the Semantic Web User Interaction (SWUI) community. Refer to the SWUI web site for current information.

Context:

  • Capturing context in a simple, usable, yet rich way -- ideally, as a by-product of other activities
  • Identifying how simple can we make it for applications to understand a user’s specific context

Process:

  • Involving the user throughout analysis and design, as well as during evaluation
  • Refining user-centered methods to support design of richer contextual/personal applications

Navigation, views, and data interpretation:

  • Scalability
  • Focusing on the relationships (the lines) in visualizations, representations, and queries
  • Maintaining ontologies -- language evolution, collaboration, relationships, rules

Authority and permission:

  • Interacting with agents -- defining your relationship with them, answering their questions, communicating changes
  • Avoiding "pestering" while maintaining oversight
  • Understanding trust, privacy

return to top of page

Questions for the usability community to consider

  • What aspects of the Semantic Web have an interface? Even if there is no interface, what are the interactions?
  • How can we be sure our interactions with data capture/management tools are so easy that anyone can -- and would -- describe themselves and the associated context?
  • How do users and technologies work together to build the semantic connections? How are they expanded and enhanced when users collaborate with each other electronically? How are old connections removed?
  • How will seeking and reasoning applications be forgiving of differences in language and meaning, and be clear/respectful of semantic ‘shades of gray’?
  • How do users remain (and feel) in control, when so much is going on ‘behind the curtain’?
  • The more that applications are built from shared, distributed components, how much harder is it to shape the overall user experience?
  • What are the fundamental implications of things like agents, reasoning engines, component technologies and web services - building blocks for web-based applications and interfaces?
  • How are user personas, preferences, experiences and interests modeled? Once these connections are built, then how are they changed as the user’s expectations and needs change?
  • How can we create better-managed semantic environments to help us avoid drowning in the ocean of metadata?
  • How does a user’s interaction with an application help filter out the myriad of possibilities, hone in on what is relevant, and make sure that the interpretations being made in the background are suitable to a particular situation that day, and not a situation faced last week or last month?
  • How much of this can be done without requiring many hours of my time informing my machines/appliances of my desires, circumstances and vocabulary preferences?
  • What interactions are required so that semantics can support online conversations and interactions? Can semantic technologies enable person-to-person and group interaction, and make the idea of online community more practical? What facilities are required to make that easier?
  • How do we keep the language from becoming stale - or, put another way, how do we maintain the Semantic Web?

return to top of page

Workshops and resources

The SWUI project page and the SWUI wiki, with discussion topics, studies, and ongoing collaborations.

Mailing list for SWUI discussion: public-semweb-ui@w3.org. (list archive is available on the web)

Semantic Web User Interaction workshop series:

Additional references:

return to top of page

Examples and Applications

The applications list was getting too out of date... so until I get it updated, I recommend going through my most recent guided tour presentation, as it contains most of the currently links that I use as a reference for my design conversations: A Glimpse into the Present of Semantic Web Interaction (June 2008).

 

return to top of page


| Welcome to IPGems | PCD & HCI | IM & KM | Contact IPGems |


© Duane Degler 2004-2010