SemanticWiki mini-series (session-1) Launch Event - Thu 23-Oct-2008    (1NS0)

Conference Call Details    (1MLB)

Attendees    (1MM4)

Session Abstract: A Survey of the Landscape and State-of-Art in Semantic Wiki    (1MMF)

In this launch session, we will ...    (1NSL)

See also: the developing SemanticWiki home page for this mini-series project.    (1MMJ)

Announcing: the Birds-of-a-Feather Meeting at ISWC (Karlsruhe, Germany) on Sunday 26 Oct 2008 (as discussed in the previous call)    (1NKS)

Agenda & Proceedings:    (1NSQ)

Presentation Material:    (1NSY)

Questions, Answers & Discourse:    (1NT5)

Questions and Discussion captured from the chat session:    (1NV7)

 PeterYim: Welcome to the SemanticWiki mini-series (session-1) Launch Event - Thu 23-Oct-2008    (1NTF)
 * Title: A Survey of the Landscape and State-of-Art in Semantic Wiki    (1NV8)
 * Session Co-chairs: Dr. SebastianSchaffert (Salzburg Research, Austria) and Mr. MaxVoelkel (FZI, Karlsruhe)    (1NV9)
 Eric Miller: I'm surprised to see the trend lines pointing down for Ontology and Semantic Web...    (1NVA)
 PeterYim: question for Max - ref. slide#17 - how is the "popularity" of, say, ontologies, arrived at?    (1NVB)
 YaronKoren: You can see at the bottom - it's just the frequency of web searches on those terms.    (1NVC)
 YaronKoren: Not all that scientific.    (1NVD)
 PeterYim: thank you, Yaron    (1NVE)
 MaxVoelkel, FZI: to slide 17: i simply queries Google Trends for search term popularity - assumption: if people google for it, it's popular    (1NVF)
 PeterYim: thank you, Max    (1NVG)
 anonymous1: are the 4th SemWiki already planned?    (1NVH)
 ChristophLange: anonymous1: we have some ideas (for ESWC 09) but not yet worked them out completely    (1NVI)
 JieBao: For 4th SemWiki, how about having it along with ISWC 2009 (Oct 2009,Washington DC)? As the 
         previous workshops are all in EU, and the US user community is growing substantially recently, 
         having a US-based workshop would actually be good. - Just for your consideration    (1NVJ)
 PeterYim: anyone who is still identified as "anonymous" ... would you kindly click on the "settings" button 
           (top center) and then swap in your name (in WikiWord format) please    (1NVK)
 ChristophLange: re semwiki4: is there potential for the n+1st ws (who would submit, and what?) 
                 -- or should the focus be broadened? Our idea was the "semantic bazaar", i.e. transitions 
                 btween formal and informal knowledge on the web (for which wikis are a good tool, but 
                 maybe not _just_ wikis)    (1NVL)
 ChristophLange: ... where "bazaar" alludes to "the cathedral and the bazaar", i.e. to agile and 
                 internetworked knowledge engineering    (1NVM)
 YaronKoren: It might be worthwhile to note the commercial semantic wikis, available only as a service 
             and not (currently) as a standalone engine - Freebase and Swirrl.    (1NVN)
 MaxVoelkel, FZI: @Yaron: Agree, Freebase and Swirrl (and Twine) are also very interesting developments!    (1NVO)
 YaronKoren: Twine is a wiki?    (1NVP)
 MaxVoelkel, FZI: well... it'S hard to say what a wiki is. not as wiki as freebase, but quite flexible 
                  and modifyable by users, too    (1NVQ)
 ChristophLange: hmm, that refers to what i meant by "not just wikis" -- it need not be called "wiki", 
                 but the way of knowledge engineering is similar    (1NVR)
 FabianHaupt: FYI: there is a twine group on semantic web:    (1NVS)
 YaronKoren: Does Twine let you edit what other people have written?    (1NVT)
 Eric Miller: Twine doesn't allow collaborative editing to my knowledge    (1NVU)
 MaxVoelkel, FZI: Is this the key of wikis? At least an important point. Is a Google Document a wiki? 
                  Is a shared network drive? ... we had long discussions on such topics    (1NVV)
 ChristophLange: a shared network drive doesn't allow for links -- that's another important wiki feature    (1NVW)
 MaxVoelkel, FZI: On windows and linux you can have links, right?    (1NVX)
 YaronKoren: I'd say a wiki has two basic requirements: collaborative editing, and preservation of version history.    (1NVY)
 MaxVoelkel, FZI: For me personally easy creation of *linked content* is a key feature    (1NVZ)
 YaronKoren: Maybe that's a requirement of a *good* wiki.    (1NW0)
 MaxVoelkel, FZI: For me, versioning is optional    (1NW1)
 YaronKoren: Hm - see slide 41.    (1NW2)
 MaxVoelkel, FZI: Versioning is really nice, of course, it relieves users from fearing about errors/refactoring    (1NW3)
 MaxVoelkel, FZI: Her eis the core community discussion about it:    (1NW4)
 HaroldSolbrig: Some of these slides seem to go beyond the classic notion of "wiki".  Is "wiki" a philosophy 
                and approach or an implementation of a broader set of principles?    (1NW5)
 MaxVoelkel, FZI: Wiki is an overloaded term: wiki engine runs a wiki site adhering to wiki philosophy 
                  - all is called just "wiki"    (1NW6)
 HaroldSolbrig: Do some of these notions of sharing encompass a distributed / federated triple-store registry?    (1NW7)
 HaroldSolbrig: Excellent point on the petri dish, imo.    (1NW8)
 Eric Miller: Re the question on the queue about tools adding semantics: 
              I'm curious, could you explain a bit more?    (1NW9)
 GaryBergCross: I would worry that letting tools add the semantics will lead to poor semantics.    (1NWA)
 YaronKoren: I'm guessing that it's a statement on the limitations of AI.    (1NWB)
 YaronKoren: But I could be wrong.    (1NWC)
 MaxVoelkel, FZI: @Gary: True, the user must understand what she is doing. 
                  Reasoning cannot be better than the del/data its based on    (1NWD)
 Eric Miller: so "tools" = automated meaning extraction?    (1NWE)
 PeterYim: Gary, the hand-button toggles your "hand raising"    (1NWF)
 MaxVoelkel, FZI: no, no *automated extraction*, but careful/meaningful interpretation. If a user 
                  adds a tag to a picture, and the tag is the name of a location, maybe ask the 
                  user if the picture is showing this location. Maybe ask the user if this 
                  assumtpion always holds.    (1NWG)
 HaroldSolbrig: Would be interested in versioning and transaction implementation...    (1NWH)
 SebastianSchaffert: is based on JTA and Java EE transactions    (1NWI)
 Eric Miller: There would need to be rules and automated reasoning tools embedded into the system, 
              though, in order to know what clarifying information was needed.  Right?    (1NWJ)
 SebastianSchaffert: Java EE and Jboss seam already provides much infrastructure for transactions    (1NWK)
 Eric Miller: (side note: I'm not a researcher, I'm a developer working on some semantically informed 
              applications, so I'm not as familiar with the state of the art here; thanks for your patience)    (1NWL)
 MaxVoelkel, FZI: Of course, meaning extraction is also a nice branch of technology, but in my view, 
                  one should ask the user, if the extraction was ok, or at least treat 
                  the two kinds of facts (explicity, extracted)    (1NWM)
 SebastianSchaffert: - SWOOKI    (1NWN)
 MaxVoelkel, FZI: Slide 51, too    (1NWO)
 MaxVoelkel, FZI: "how to do as much semantics as possible with as little user exposure as possible"    (1NWP)
 YaronKoren: Well, this "challenge" ("if it is not useful in semantic wikis, it is not useful at all") 
             seems like a bit of a tautology - given that any kind of content can show up on a wiki page, 
             for something to be useful there seems to just mean that it's useful.    (1NWQ)
 ChristophLange: little user exposure could generally mean: whenever the user contributes sth., 
                 assume that (s)he is willing to give more information in that moment. So ask him/her about 
                 why he did so, and record the answer in the knowledge base (as in Max's example on photo tagging)    (1NWR)
 HaroldSolbrig: I'd like to argue that there are two functions - the creation of ontological knowledge 
                (wikipedia, semantic wiki, tagging, etc) and knowledge maintenance - Protege and like tools 
                that are used by specialists to integrate and coordinate.    (1NWS)
 SebastianSchaffert: it is like software engineering    (1NWT)
 SebastianSchaffert: much simpler now, because experts become quicker    (1NWU)
 HaroldSolbrig: We use a model where SME's suggest in whatever degree of formalism they are comfortable, 
                suggestions are exported, Ontology engineers integrate with specialized tools, return to 
                the semantic wiki for sign-off or correction.    (1NWV)
 HaroldSolbrig: We've lost too much SME value by making them fit their knowledge to the tool rather than v-v, 
                but we still need quality and consistency in the output.    (1NWW)
 HaroldSolbrig: Will present on this in January.    (1NWX)
 SebastianSchaffert: and also there are design principles where I don't need an engineer to build simple software, 
                     our web designer can do that    (1NWY)
 PeterDolog: Harold, I'm looking forward to hear on that in january    (1NWZ)
 PeterDolog: it seems like a pattern    (1NX0)
 HaroldSolbrig: You touched on that in the talk.  The notion that Wikis are organic through and through.  
                We've encountered some real issues when jit software collides with traditional software 
                development methodologies.    (1NX1)
 HaroldSolbrig: SW development is going to have to change - is a template software or content?  How about a query?    (1NX2)
 HaroldSolbrig: How about when a user tweaks js or css    (1NX3)
 YaronKoren: These are all examples of scripting languages.    (1NX4)
 HaroldSolbrig: Indeed, but now the users can treat them as data, vs. requirements -> devel -> test -> deploy.    (1NX5)
 Eric Miller: "when jit software collides with traditional software development methodologies": could you 
              expand on that?    (1NX6)
 HaroldSolbrig: I can't go into, say, Amazon and change their pages today.    (1NX7)
 HaroldSolbrig: Part of what makes Wiki and Semantic Wiki powerful for us is that it is easy to change 
                templates, scripts, plugins, etc. to customize the environment to specific SME requirements. 
                (Semantic Forms & related tools as well).    (1NX8)
 HaroldSolbrig: If the environment doesn't fit your needs, we tweak the forms and associated tools so they do.    (1NX9)
 HaroldSolbrig: If I change a form, does it have to go through sw change control?  (We are hosted on a 
                government site, which makes  it even more tricky - the EPA got burned really bad by some 
                unsolicited content...)    (1NXA)
 HaroldSolbrig: Excellent presentations!!    (1NXB)
 PeterDolog: bye to all    (1NXC)
 SebastianSchaffert: bye    (1NXD)
 FabianHaupt: cya all    (1NXE)
 FabianHaupt: happy coding    (1NXF)
 MaxVoelkel, FZI: Thanks for being a great audience!    (1NXG)
 ChristophLange: - please consider joining if you are 
                 there, Sunday 6pm    (1NXH)
 MaxVoelkel, FZI: I'll be there, too!    (1NXI)
 ChristophLange: sure    (1NXJ)
 HaroldSolbrig: Wish I could - will any of it be webcast?    (1NXK)
 PeterYim: see you there on Sunday, I'm coming too!    (1NXL)
 ChristophLange: technically not possible, I fear. But the slides of the talks will be online afterwards    (1NXM)
 PeterYim: thanks everyone, for your participation and contribution!    (1NXN)
 ChristophLange: thanks for organizing!    (1NXO)
 PeterYim: Christoph ... thank you very much for the synthesis and the compilation of the "hot topics" 
           from the 9/18 session ... that was most wonderful!    (1NXP)
 YaronKoren: Maybe it makes sense to turn the December session into just a Semantic MediaWiki session?    (1NXQ)
 YaronKoren: It might be too many topics to cover otherwise.    (1NXR)
 PeterYim: we just don't seem to have enough sessions! ... I agree, Yaron, so much to cover just in SMW 
           and its extensions    (1NXS)
 ChristophLange: @Peter: thanks for the compliment, pleased to hear that you liked it    (1NXT)
 PeterYim: maybe after the first six ... we could delve deeper into particular topics (like SMW, using 
           SWiki for ontology repositories & registries, etc.)    (1NXU)
 MaxVoelkel, FZI: A round presentation about SMW + the ecosystem of extensions would be great. I think 
                  this does not exist yet in any concise form.    (1NXV)
 PeterYim: let's all keep the suggestions and discussion going on the [swikig] list ...    (1NXW)
 YaronKoren: Is that where these sessions were planned? I was wondering about that.    (1NXX)
 MaxVoelkel, FZI: Session planning happened partially in the ontolog wiki, partially between PeterYim, 
                  Sebastan and myself, HaroldSolbrig and LiDing too, partially in the last call, ...    (1NXY)
 YaronKoren: Ah.    (1NXZ)
 MaxVoelkel, FZI: Session 2-6 are not cast in stone, but seemed to make a nice structure to cover most topics    (1NY0)
 PeterYim: it was supposed to ... we have a "program committee" of sorts  (we refer to them as co-champions here) 
           who was supposed to seed that discussion ... todays' presented "plan" (which is supposed to be a 
           strawman) is the first time this has been "socialized"    (1NY1)
 PeterYim: the program needs to be community driven, and done in the "wiki way!"    (1NY2)
 YaronKoren: Well, where's the page where it's being discussed?    (1NY3)
 MaxVoelkel, FZI:    (1NY4)
 PeterYim: I leave for Germany at about 5am tomorrow ... if I have time (after taking care of urgent matters) 
           I'll work on today's audio archive and post it before my trip) ... if not, I'll work on that 
           after I get back from ISWC (Karlsruhe) and FOIS (Saarbrueken)    (1NY5)
 MaxVoelkel, FZI: Have a safe trip & cu here!    (1NY6)
 PeterYim: this mini-series project home page is at: 
           ... one can start from there    (1NY7)
 YaronKoren: How can I start?    (1NY8)
 PeterYim: I mean start reading ...    (1NY9)
 YaronKoren: Oh.    (1NYA)
 PeterYim: Thanks ... see you (going to lunch now)!  bibi    (1NYB)
 ChristophLange: CU on Sunday, have a good trip!    (1NYC)
 PeterYim: Tx ... bibi (for now)    (1NYD)
 ChristophLange: bye!    (1NYE)

Session ended 2008.10.23 12:20 pm PDT    (1NTH)

Audio Recording of this Session    (1NTI)