Home > Semantic Web > Human and the Semantic Web

Human and the Semantic Web

July 16th, 2008

“The Semantic Web is mainly serving machine agents” has been dominating my mind for many years. Now human users may also want to explore the big mass of RDF data not just for debugging purpose. Semantic Web user interaction is becoming an important part of Semantic Web layer cake and research direction (see SWUI workshops) in ISWC.

As a “web of data”, the Semantic Web, boosted by Linked Data efforts, presents web users a maze of RDF graph with billions of arcs (triples). To explore the maze, below are some html browser approaches I came across:

An alternative approach is graphical browser, which seem to be more intuitive to end users. An interesting blog Large-scale RDF Graph Visualization Tools covered a handful of useful resources including something I never encountered and even links to 28 visualization software packages. Of course the list missed some RDF viz browsers such as FOAFnaut, Welkin, and self visualization. It is notable that scalability is still bugging most of the visualization approaches due to the limit of memory size: my last experience was “Otter had a hard time when processing a graph with over 10,000 nodes”.

There are still many user interaction issues beyond the browsers (e.g. search engines, semantic wiki), and a well-designed UI component is probably the key to the Killer-App of the Semantic Web.

Li Ding

VN:F [1.9.22_1171]
Rating: 0.0/10 (0 votes cast)
VN:F [1.9.22_1171]
Rating: 0 (from 0 votes)
Author: Categories: Semantic Web Tags:
  1. July 17th, 2008 at 03:21 | #1

    At Swirrl, we are trying to overcome some of those hurdles around humans viewing RDF information.

    These articles on our blog help explain how we’re going about it.

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  2. July 17th, 2008 at 04:43 | #2

    Hi Li,

    Nice blog post. You are very right, now that we have the data people have started to turn their attention to things like:
    * More reliable trust solutions
    * Proof/Rules
    * User Interfaces

    It is particularly the User Interfaces being worked on which I am quite glad to see, because traditionally user interfaces have been pretty poor for Semantic Web applications… but this is obviously changing.

    For example, at OpenLink we’ve evolved OpenLink RDF Browser into the OpenLink Data Explorer, as it is no longer just about browsing through exposed RDF data because we can now sponge things efficiently on the fly. Our OpenLink Data Explorer provides a layer between the data and the user’s browser, and so it’s the integration with the browser which provides some useful user-friendly features. There’s a lot more to it than I’ll go into here, but you can find out more: http://ode.openlinksw.com/

    This is just the beginning of a more user-friendly and useful web.

    I’d like also like to add:
    * More Hyperlink based Linked Data Viewers: Marbles, Zitgist Data Viewer and Triplr
    * Another Semantic Web search engine: Syndice
    * A nice and flashy ontology system: Umbel

    I do urge you to have a look at the ODE Browser Extension.

    It brings me on to something else which is the Semantic Web Layer Cake, which I don’t think exists. I think there would be a lot more benefit if we look at things in a kind of component based plug and play type model (I explained this in my blog post “WUPnP“)

    Cheers,

    Daniel

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  3. July 17th, 2008 at 08:51 | #3

    Hi Daniel,
    thanks for sharing a lot more useful resources. I tried to used ODE browser extension. It offers a lot of interesting and useful features. There are still many issue need our answers:
    * what is the minimal knowledge before normal web users can use these tools?
    * Should we simply hide RDF and URIs from the end users unless they want to “view source” ?
    * Is hyperlink enough for users to browse the (Linked) semantic web?

    best,
    Li

    VN:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VN:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  4. July 17th, 2008 at 12:03 | #4

    Li,

    Re. RDF Interfaces, I don’t think the problem is solely user competence level. In my 20+ years of experience with data and user interaction, I know that “users” are dumb. Of course, they may be disinterested, but this doesn’t make them dumb or capability challenged.

    On the developer side of the fence, we sometimes try to “over dumb down” interfaces to our work without undersatnding that “User Interaction” is just that “User Interaction” which means we have to understand protocols that experss a degree of politness, which then stimulate the users attention.

    The Web is primarily experienced via Web Browsers. These ubiquitous user agents have always offered:
    1. Ability to view rendered markup
    2. Ability to view the markup

    Now I know for sure that most end-users aren’t interested enough to grok the markup exposed via the typical “View | Page Source” menu or context menu interaction sequences, but when they do, the protocol (politeness of the feature exposed via menu etc.) lays the context for the interaction.

    The approach we’ve taken with the OpenLink Data Explorer (a derivative of the initial OpenLink RDF Browser) is the simple additon of a 3rd feature for interacting with Web Pages via the “View | Linked Data Sources” main or context menu navigation sequence. We’ve politely added a new option to the browsing experience that does the following:

    1. Let’s the user know they are switching browser view modes
    2. They are seeking to view the raw data sources associated with a Web page

    (imho), this about the adition of an optional feature that is polite enough to stimulate the mental transition of the user from “viewing a Web Page” to looking at what a Web page is about via its associated Linked Data Sources.

    The Linked Data we expose is the result of RDFizing the source (X)HTML URI via server side cartridges bound to a Web Service (that we call the Sponger).

    Links:
    1. https://addons.mozilla.org/en-US/firefox/addon/8062
    2. http://ode.openlinksw.com
    3. http://ode.openlinksw.com/example.html

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  5. Jie Bao
    July 18th, 2008 at 12:03 | #5

    Hi Li

    I don’t believe browsing RDF will be a big thing for end users anyway. In fact, a good SW application should hide RDF as much as possible. How many of Hotwire or Expedia users are aware of relational databases the sites are relying on? One key of success of a database application is to make information easier to input and easier to read – and to make the information making sense to end users. It might be the same for SW applications.

    I would ask myself the following questions whenever I develop something that is claimed to an “application”: if the information offered by the application makes sense? how many minutes a typical user need to learn my application for doing something very simple? If after 5 minutes, a typical user still has no clue to use or understand the application, I would say the application is failed (as an end-user application). Unfortunately, it is the very reason why so many SW applications fail to attract users.

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  6. Tim Lebo
    July 18th, 2008 at 18:08 | #6

    Perhaps we could leverage the flexibility and power of RDF to visualize it? I thought the Fresnel Display vocabulary was a good start — using RDF to specify what triples to show and how to show them.

    From what I have been able to discern, the intersection between UI and semweb has been skipping along for the past few years (as Li Ding points to above). I think the hurdle is to convince the information visualization folks that the semweb ecosystem products can help them do what they like doing: visual communication of complex data sets. One chore they consistently need to deal with is manipulating Yet Another Format.

    I’ve started a couple of twines in hopes of aggregating the discussions and technologies.

    * http://www.w3.org/2005/04/fresnel-info/
    * http://www.twine.com/twine/116nlm5b5-c21/semantic-web-user-interaction
    * http://www.twine.com/twine/116nk71gv-bfr/visualization-ontology

    Regards,
    Tim

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  7. August 3rd, 2008 at 15:20 | #7

    That was a well written article, very intereting,thank you for a good read.

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  8. April 15th, 2010 at 01:45 | #8

    There is an Ontogloss, ontology based annotation tool that uses pre–defined concepts in ontology to mark–up a document. Just share!
    http://emeld.org/workshop/2005/papers/mostowfi-paper.html

    VA:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
    VA:F [1.9.22_1171]
    Rating: 0 (from 0 votes)
  1. August 3rd, 2008 at 23:23 | #1