ppy/EarthScienceOntolog-s02_chat-transcript_edited_20120906b.txt --------- Chat transcript from room: ontolog_20120906 2012-09-06 GMT-08:00 [ time-stamps in PDT ] --------- [08:47] PeterYim: Welcome to the = EarthScienceOntolog: Panel Session-02 - Thu 2012-09-06 = Mini-Series Theme: An Earth Science Ontology Dialog ("EarthScienceOntolog") Session Topic: Ontology Development and Application across Earth Science Systems Lifecycle Session Co-chairs: Dr. GaryBergCross, Dr. NaicongLi Panel-Briefings: * Dr. GaryBergCross (SOCoP; Knowledge Strategies) - EarthScienceOntolog Session-2 Intro: "Understanding Semantic Insertion within the Earth Science Application Lifecycle" * Dr. RichardHooper (CUASHI) - "Implementing a community-governed ontology: Experiences from the Water Sciences Community" * Professor PeterFox (RPI/TWC) - "Life cycle semantics in Earth and space sciences - what's worked (and not) and where are we... a decade in..." * Mr. MattiaSantoro (CNR, Italy) "A Semantic Broker for enhancing resources discovery," a work by StefanoNativi, MattiaSantoro, C. Fugazza, M. Craglia * Dr. NaicongLi (U of Redlands), Dr. PhilipMurphy (U of Redlands), Professor KrzysztofJanowicz (UC Santa Barbara) - "An ontology-driven framework and web portal for spatial decision support" Logistics: * Refer to details on session page at: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2012_09_06 * (if you haven't already done so) please click on "settings" (top center) and morph from "anonymous" to your RealName * Mute control: *7 to un-mute ... *6 to mute * Can't find Skype Dial pad? ** for Windows Skype users: Can't find Skype Dial pad? ... it's under the "Call" dropdown menu as "Show Dial pad" ** for Linux Skype users: please stay with (or downgrade to) Skype version 2.x for now (as a Dial pad seems to be missing on Linux-based Skype v4.x for skype-calls.) --- On the chat: AlexSorokine, AnneThessen, BobbinTeegarden, Chuck Ward, ClarePaul, DeanaPennington, DaliaVaranka, DeborahNichols, DonElsborg, DougFils, DougFoxvog, ElizabethFlorescu, FrankChum, FrankDAgnese, FrankOlken, Francois de Beuvron, GaryBergCross, GenhanChen, JackRing, JoelSachs, Kai Liu-GMU, KenBaclawski, KrzysztofJanowicz, LeoObrst, MarcioFaerman, MarshallXMa, MattiaSantoro, NaicongLi, NancyWiegand, PatriceSeyed, PeterFox, PeterYim, PhilipMurphy, Rich Keller, RickHooper, Robert Downs, RuthDuerr, Scott Hills, ScottPeckham, SiriJodhaKhalsa, SumitPurohit, TaraAthan, TerryLongstreth, TomTinsley, Uma, --- . == Proceedings: == . [08:50] anonymous morphed into MattiaSantoro [09:18] anonymous morphed into TomTinsley [09:22] anonymous morphed into NaicongLi [09:23] anonymous1 morphed into GenhanChen [09:25] anonymous1 morphed into RickHooper [09:25] anonymous morphed into DonElsborg [09:26] anonymous morphed into PeterFox [09:26] GaryBergCross: Hello Peter et al [09:27] GaryBergCross: I see that all the speakers are on the chat! [09:28] anonymous morphed into DougFils [09:28] anonymous1 morphed into ElizabethFlorescu [09:29] anonymous3 morphed into ClarePaul [09:29] anonymous1 morphed into KaiLiu-GMU [09:30] anonymous morphed into DeborahNichols [09:31] SiriJodhaKhalsa: is there any way to preserve the chat session? [09:50] LeoObrst: @SiriJodhaKhalsa: Yes, the chat will be preserved. Ontolog always captures these, along with an audio file of the presentations. [09:31] anonymous1 morphed into PhilipMurphy [09:32] anonymous1 morphed into PatriceSeyed [09:33] anonymous1 morphed into RuthDuerr [09:33] anonymous morphed into FrancoisDeBeuvron [09:35] PeterYim: == GaryBergCross presenting the introduction to this session ... [09:36] anonymous morphed into DougFoxvog [09:36] anonymous1 morphed into FrankOlken [09:37] anonymous1 morphed into PhilipMurphy [09:37] anonymous morphed into AlexSorokine [09:38] anonymous1 morphed into MarshallXMa [09:38] anonymous morphed into JoelSachs [09:38] SiriJodhaKhalsa: the diagram was a consensus diagram developed at the EarthCube PI workshop. It originated at CNR [09:39] SiriJodhaKhalsa: this diagram (slide 5) did originate w/ Interop group. the previous consensus diagram originated with the brokering group [09:47] GaryBergCross: @SiriJodhaKhalsa Thanks for the history of the graphics. I should have said that I got them from INTEROP Concept group slides. [09:42] anonymous morphed into TaraAthan [09:47] PeterYim: @Gary - in between speakers, would you please remind those on the call, who haven't logged into the chat-room, to join us there (we now have 48 people on the conference bridge, but only 37 in the chat-room [09:42] PeterYim: == RickHooper presenting his talk ... [09:51] SiriJodhaKhalsa: why distinguish between ET from atmos energy budget and a pan measurement? [09:51] SiriJodhaKhalsa: conceptually? [09:55] SiriJodhaKhalsa: ScottPeckham - you've given a lot of thought to what should be in a variable name, in the metadata and in an associated ontology. seems like these are getting mixed up [10:01] ScottPeckham: Hi SiriJodhaKhalsa. The issues with data discovery and knowledge capture are different than the "semantic mediation" issues that the CSDMS Standard Names were designed to address. [10:03] ScottPeckham: The CSDMS Standard Names are described on our wiki at: http://csdms.colorado.edu/wiki/CSDMS_Standard_Names [09:56] JackRing: "When should scientist consult original source?" Always, and first. The use of surrogate terms to represent context ALWAYS yields unacceptable Type 1 and Type 11 errors. Pls be aware that you are being influenced by the presumption of von Neumann machines whereas modern data relationship/flow machines are now available. [09:58] SiriJodhaKhalsa: Mattia - convey across communities or simply provide brokering service? [10:04] MattiaSantoro: @SiriJodhaKhalsa - I think the two things are not in "contrast", brokering is a way of homogenizing metadata and data models across communities and information systems [09:59] JackRing: TKU, Rick. Well said. [10:02] RickHooper: There is a cost to documenting data for re-use that data publishers tend not to want to incur. It doesn't help them personally. We need to reward the effort required for data documentation. [10:07] FrankOlken: Rick, How are you storing (and retrieving) the various hydrological time series: a relational DBMS, HDF5 files, ??? [10:07] RickHooper: relational DBMS [10:48] FrankOlken: Rick, Out of curiosity, which relational DBMS are you using for the hydrological time series? [10:55] RickHooper: @FrankOlken: This is all on the MS platform, so we are using MS SQL Server [10:00] PhilipMurphy: what was the URL to the HIS project? [ http://www.cuahsi.org/ ... --ppy/added subsequently ] [10:00] PeterYim: == PeterFox delivering his brief ... [10:00] anonymous morphed into RichKeller [10:00] anonymous1 morphed into RobertDowns [10:06] anonymous morphed into SumitPurohit [10:11] MarshallXMa: http://dx.doi.org/10.1007/11926078_57 and http://dx.doi.org/10.1016/j.cageo.2007.12.019 are links to the two papers Peter mentioned. With some details of the ontology works [10:34] GenhanChen: Rick, on your slide 6 and 7, for the hierarchical data, do you just map "leaves" to variables for search? I am wondering how you deal with their parents' information (e.g. core concept, property or branch). [10:38] RickHooper: Genhan: Yes, but I should note that a leaf can be attached to more than one branch, so it is not a strict hierarchy. There is no user input to the hierarchy--we control it. Does that answer your question? [10:41] GenhanChen: Rick, does it mean the parents' information will be lost? [10:55] RickHooper: @Genhan: Not sure what you mean. The parent information is not 'lost', but it is not transmitted to the user with the data or metadata return. Only the concept and the publisher's variable name is transmitted to the user. [11:03] GenhanChen: Thank you, Rick. That is what I would like to know. [10:25] PeterYim: == MattiaSantoro presenting ... [10:31] JackRing: What are the limitations of searching words in free text? [10:36] SiriJodhaKhalsa: automatic mapping is the crux, and most challenging aspect of 3-party mediation [10:43] SiriJodhaKhalsa: question for Mattia: will the semantic broker be able to discover and utilize vocabularies/thesauri/ontologies associated with a data source? [10:46] SiriJodhaKhalsa: Mattia - does a new adapter have to be created for each new semantic resource? [10:54] MattiaSantoro: @SiriJodhaKhalsa - A new adapter has to be created for each communication protocol and/or data/metadata model [10:46] JackRing: Mattia. Will you be measuring False Positives and False Negatives? [10:56] MattiaSantoro: @JackRing - yes false positive is one thing we will measure to improve the association strategy [10:59] MattiaSantoro: @JackRing - the main issue with free text is in multidisciplinary contexts, where the same word might have different meanings depending on the discipline (or sometimes even depending on the language) [11:05] JackRing: Mattia, TKU. How will you measure false negatives, i.e., things that existed but not found? [11:07] JackRing: Mattia, Cannot the ambiguity of what a word signifies be resolved by assessing context? Are you seeking to avoid combinatorial explosions in the computations? [10:47] PeterYim: == NaicongLi presenting ... [11:03] FrankChum: Great presentation, NaicongLi. Sorry I have to jump off to another meeting. [11:09] PeterYim: == open Q&A and discussion ... please click on "hand" button (lower right) to notify the chair, and when recognized, announce yourself and make sure you can be heard before speaking [11:09] anonymous1 morphed into FrankDAgnese [11:14] BobbinTeegarden: @all speakers What are you using for visualization, and is there any direct interaction with the visualization itself? [11:14] RickHooper: Startree viewer is what we are using, but there are problems with that. [11:14] KrzysztofJanowicz: [ref. verbal question from SiriJodhaKhalsa] Are you asking about ontology alignment? [11:17] GaryBergCross: >[Krzyszto] Are you asking about ontology alignment? - I don't think so. [11:16] SiriJodhaKhalsa: thanks [11:19] KrzysztofJanowicz: @SiriJodhaKhalsa: With respect to Spatial Data infrastructures there is work on the automatic meditation of sensor metadata and observation data to ease the registration at Sensor Observation Services. I would consider this similar to a brokering with a reduced amount of human interaction. [11:23] anonymous morphed into DaliaVaranka [11:23] JackRing: Whereas each of us has a 25.000 to 75,000 word vocabulary many projects/companies run just fine with 2500 word vocabularies. Local ontologies are necessary. However Log(variety) = K*Log(extent) so as we seek interoperability among ever-larger communities then the ability to interoperate local ontologies becomes valuable. [11:29] MarcioFaerman: @JackRing, could you please explain what you mean by K, variety and extent in your previous statement? Thanks! [11:33] JackRing: @Marcio. As a system gets larger in scope (extent) it inexorably gets increasingly heterogeneous (variety of elements and relationships). In complex systems research we often not that the relation is log(variety) = 0.7*log(extent) however many systems are messy so we even see that variety increases in a static system. [11:34] MarcioFaerman: Tks @JackRing! [11:25] KrzysztofJanowicz: [ref. RickHooper's remark that "some taxonomy is important, but it is not important to achieve a consensus on the taxonomy" (as different people use different logics to organize their terms, and their taxonomies will come out different)] I could not agree more. [11:28] KrzysztofJanowicz: I agree that while terminologies are important, common agreement is not always important and difficult to archive. [11:28] Gary Berg-Cross: Krzysztof comment above was in the context of my question on value of taxonomies and modular ontologies. [11:29] GaryBergCross: I want mention the Dayton Vocamp this Sept 15-17 which may be of interest to some session attendees. Kno.e.sis Center at the Department of Computer Science and Engineering of Wright State University in Dayton, OH focus on the observation-driven engineering of geo-ontologies by developing small, self-contained, and reusable geo-ontology design patterns that can be used to annotate geo-data http://vocamp.org/wiki/GeoVoCampDayton2012 [11:32] PeterYim: == GaryBergCross wrapping up the session ... [11:32] PeterYim: join us again for session-3 of this mini-series on Oct-11 ... please mark your calendars [11:32] PhilipMurphy: Could we have such a Use Case <> Ontology development Ontolog session in the future? [11:33] GaryBergCross: @PhilipMurphy - I'm all for it [11:33] MarcioFaerman: Ditto, @PhilipMurphy [11:34] Scott Hills: Re: Future Use Case <> Ontology session - I agree. [11:33] PeterYim: [added subsequently] @PhilipMurphy & All - Indeed! Also, the 5th session of this mini-series slated for Dec-6 entitled: "Tutorial" to be chaired by NancyWiegand and MikeDean, could well be part of what you are looking for. Come join us then, and let's try to refine on what kind of session(s) would be most useful. [11:33] PeterYim: great session, thank you ALL, speakers and participants! [11:34] PeterYim: -- session ended: 11:34am PDT -- ---------