Actions

Ontolog Forum

A Glimpse into the Future of SemanticWiki's

For SemanticWiki mini-series Session-6: The Future of Semantic http://c2.com/cgi/wiki? Trends, Challenges and Outlook (Panel Discussion)

As part of the preparation for the concluding session of the SemanticWiki mini-series program, we are inviting all our panelists to share with us their vision of what the future of semantic wikis would be like.

All organizers, chairs and speakers from previous sessions are invited to each deliver short statements regarding the future of semantic wikis as they each see it (2 min. each). Those who plan to make short statements are requested to add their 3 to 6 bullets here (or send them in via email to the session co-chairs with a copy to <peter.yim@cim3.com>) before the end-of-day Friday 2009.02.27

Key bullets of speakers' short statements

Sample statements: (template)

    • 1. the term "semantic wiki" will disappear as users adopt it into mainstream wiki applications
    • 2. semantic wikis will be one of the most important vehicle in bringing the semantic web to the masses
    • 3. by 2012, over 40% of open virtual communities, and over 25% of corporate intranets will be

utilizing wikis; and half of them will be (what we would recognize as) semantic wikis.

Please enumerate your 3 to 6 thoughts on how you see the Future of Semantic Wikis, in terms of Trends, Challenges and Outlook.

... read this if it's your first time editing the Ontolog wiki. =ppy

These thoughts will be presented in 2-minutes lightning talks at the SemanticWiki mini-series session-6 event

  • Christoph Lange
    • technologies from semantic wikis will merge into collaborative Web 2.0 (3.0?) apps like Google Docs
    • nobody will use wiki text syntax any more. WYSIWYG HTML editing or semantic forms or something else will be the GUI for editing/annotating
    • still a challenge: adequately mapping knowledge with complex structures onto the page-granular wiki architecture, where the page is the base unit of editing, versioning, linking, search
    • still a challenge (not really been addressed in recent years AFAIK): more expressive reasoning vs. maintaining consistency
  • Daniel Hansch
    • Semantic wikis provide an enormous potential to serve communities of casual users as mash up application providing unified view on business-related artifacts. We foresee the following challenges on the way to this vision:
      • Improve usability of schema level operations (e.g. wizzard to build a representation of the artifact "customer complaint")
      • Improve ways to include artifacts from legacy systems into the wiki (e.g. drag and drop of documents, bugzilla bug reports etc)
      • Learn about usability and ease-of-use from commercial (enterprise) wikis such as Confluence or Deki Wiki
  • Daniel Schwabe
    • Semantic Wikis are a special case of the more general category of Model-driven Web Applications, for which established methodologies exist;
    • Future (semantic) web applications will have wiki characteristics, notably, end-user contributed data AND schemas;
    • The current (semantic) wikis architectures are not appropriate to support full-fledged applications; more general architectures will be used.
  • Harold Solbrig
    • "Ontology" as it is used today serves (at least) two distinct roles:
    • 1. A collection of shared definitions:
    • --- When I (we) say "X", we mean...
    • --- If you've got A, B, C then you've got an X
    • 2. A distributed collection of knowledge
    • --- X might cause Y.
    • --- X usually is associated with T
    • --- ...
    • A primary role of Wiki (however we define it) has always been role 2
    • Semantic Wiki allows islands of knowledge to become networks of knowledge, by allowing links across definitions... if we have shared knowledge of those definitions. (O4W)
    • The jury is still out when it comes to the applicability of Semantic Wiki on role (1). We have seem some promising efforts on W4O, and, clearly, Semantic Wiki can play a critical role in evaluating, critiquing and gathering collective knowledge for role 1, but the OOR still seems to be something that goes well beyond "wiki" space when it comes to organizing, validating and dissemination.
    • Application and solutions are still fragmented - we need more mix and match, both on content AND tools
  • JoelNatividad
    • We're in the [[WordStar/WordPerfect]] stage
    • Successful Semantic Wikis require Knowledge Curators/Evangelists
    • Auto-annotation may help win the argument against "boiling the ocean"
    • Create magnet content/applications that make it worth it
    • Jumpstart Semantic Wikis by pumping legacy data into them
  • Kei Cheung
    • 1. Community buy-in: we need to convince the users (e.g., neuroscientists and life scientists) that semantic wiki can help their work
    • 2. If we have DBPedia as the semantic web interface to Wikipedia, how can we create a more powerful interface between semantic web and semantic wiki?
    • 3. Can semantic wiki help bridge web 2.0 and web 3.0?
  • Mark Greaves - Trends and Outlooks: Three Areas to Watch
    • User Interfaces are THE problem
      • Semantic Wikis are tools for Publication and Consensus around semantic data
      • Wikis are now an IT code word for "zero-training"
      • Semantic-specific wiki UIs must dovetail with the "wiki way"
    • We haven't yet retired the Knowledge Engineer
      • There is no persuasive evidence that "schema-last" social systems like semantic wikis will converge on good, usable ontologies
        • We need to find the places where the crowd will converge on something valuable
      • Currently, good semantic authoring still requires some amount of gardening
        • Authoring tool support with wizards, check-bots, etc
        • Knowledge engineers actually know something
      • How do we achieve network effects in semantics?
    • Resist Scope Creep
      • Semantic Wikis appear to have a "sweet spot" in RDF(S)-style KR
      • Other tools can supplement wikis when the KR requirements get too complex
  • Mark Musen - regarding "what is a semantic wiki?"
    • [One possible issue] is that there are folks who are ascribing a religious aura to the term "semantic wiki." People in the community are focusing on the "wiki" moniker, rather than on the underlying functionality that has been implemented. It has become an article of faith that, if developers are using the term "wiki", then the software is, ipso facto, easy to use. If they are using the term "semantic wiki", then not only is the software easy to use, but also by definition the software "solves" the problems of knowledge representation. I have heard of representatives of funding agencies that are uninterested in any Web 2.0 technology unless it is a "wiki", since "wiki" has become synonymous with usability.
    • I am not suggesting that the "semantic wiki" community needs to draw lines in the sand, but I am concerned that the term "semantic wiki" is becoming an overloaded buzzword that often drives unrealistic expectations. I agree that the community should be inclusive, but I wonder whether the term "wiki" is no longer helpful. I'm having a lot of déjà vu from the days when the phrase "expert system" became so problematic for the knowledge-based systems community.
  • Markus Kroetzsch
    • Modeling and using higher-level knowledge will be more important

(application-specific data structures, reasoning, constraint checking, etc.)

    • Exchange of information between servers will increase (wiki-to-wiki, but

also between other sites/tools/web services)

    • Wikis (semantic or otherwise) will vanish as a separate type of application
  • Mike Dean
    • Semantic Wikis (particularly SMW) need to support more of OWL. Domain and range are insufficient.
  • Peter Dolog
    • semantic wikis as user programmed environments for applications (business, public and other domains)
    • programmed in the same way as the wiki pages are edited
    • different models and languages need to be studied to support that (such as composition by meshup, lightweight domain specific languages and so on)
  • Peter P. Yim
    • The semantic wiki (in its enlarged scope) as the HCI (human-computer interaction) interface of choice
    • Wikis, Semantic Wiki and even the Semanitc Web to disappear into the mainstream, to be recognized by users and the media as web x.0
    • Impact: Semantic Wiki as the enabler of collective human intelligence - I am constantly reminded of Doug Engelbart's NLS (oNLine System), and his notions of OHS (Open Hyperdocument System), DKR (Dynamic Knowledge Repository) and [[CollectiveIQ]] . . . finally, after decades, they are coming close to becoming a reality
    • a word of caution: if we give up on "open" - open read/write access, open content, open source technology; essentially open collaboration & open knowledge - as many applications may opt for, we are giving up on a huge portion of the power and benefit that a wiki (as it was originally conceived) is able to bring us.
  • Rudi Studer - Some Decisions to Be Made (by Us)
    • "More Power!" vs. "Keep it simple!" (a classical conflict, continued)
      • SemanticWiki users do ask for more expressivity
      • SemanticWiki users do complain about ease of use
      • Semantic Tech offers so many new features
      • Approach 1: Invent new interface metaphors (this is more than just implementing impressive GUIs; e.g. use "concepts", "rules", "inheritance" not "logical formulas", "axioms", "ontologies"; new ideas wanted!)
      • Approach 2: Separate concerns in modeling (not all users are equal: clarify different editing roles; provide filtered "views" on knowledge; modularize)
      • Further Approaches?
    • What about the world beyond wikis? (yes, there is one)
      • Most SemanticWiki's offer some data exchange
      • Few SemanticWiki's actually exchange any data
      • Semantics should enable knowledge exchange
      • Approach 1: Conquer the Desktop (much data still is offline; build connections to exchange knowledge with relevant tools; often tool-by-tool basis)
      • Approach 2: Conquer the Web (every wiki is an island; how to integrate data across sites? Integrating wikis with other web services?)
      • Further Approaches?
    • What becomes of the "Wiki Way"? (times are a-changin' ...)
      • Wikis were "defined" as CMS that are quick, simple, collaborative but hardly any concrete feature is common to all "wikis"
      • Semantic Wikis are Wikis that "use" semantics
      • "Definitions" of other CMS types hardly better
      • Are these categories really useful?
      • What content management do we want to use tomorrow?
      • Semantic wikis as the basis for guiding new systems for managing knowledge
  • Sebastian Schaffert
    • Semantic Wikis as a Platform!
      • media industry really needs by now platforms that combine social software, social networking and semantic technologies
      • Semantic Wiki idea is applicable to a large set of Semantic Social Software applications
      • core technology is always the same
    • Semantic Wikis as a seed for Semantic Social Software
  • TobiasKuhn
    • I see two key requirements for the future break-through of semantic wikis:
      • The semantic representations within the wikis must give *immediate* benefit.
      • The interface of the wikis must be so easy to use that an average web user

is able to access and understand it immediately, and is able to learn how the content can be extended and changed within very short time.

    • My prediction: Many web applications will appear in the future that use

semantic technologies and that follow the principles of wikis. They will build upon the research we are doing today on semantic wikis. However, these web applications will probably not call themselves "semantic wikis".

  • Yaron Koren
    • For human-generated information, semantic wikis will become the dominant medium.
    • For machine- and externally-generated information, semantic wikis will serve as the "glue" bringing together different information sets (see, e.g., http://opencongress.org/wiki - the shape of things to come).
    • Semantic wikis will take many forms, collaborative spreadsheets being a key one.

Additional pertinent thoughts: (all are welcome!)

... please enter your thougths below (kindly identify yourself and date your entry.)

  • ... (name/date)
    • ... (remarks)