The Tetherless World Weblog

AGU experiences 2010

This is at least my fifth trip to San Francisco for the fall meeting of the American Geophysical Union, and by far and away, this one was the absolute best experience I’ve ever had. From arrival to departure. The weather was great, the company was great, the posters were great, the sessions were great, the talks were great, the meetings were great, and the social events were great.

Attending from the Tetherless World Constellation were professors Peter Fox and Deborah McGuinness; software engineers and researchers Stephan Zednik, Cynthia Chang, and myself; and graduate students Evan Patton and Eric Rozell.

As with any conference like this, we took some time to meet with colleagues and collaborators. We were able to meet with the folks at Woods Hole Oceanographic Institution, mainly on Monday morning, but also throughout the week. We also had meetings with the folks from Goddard Space Flight Center and folks from the University of Texas El Paso Cybershare group. We made a lot of good new connections and built up some possible collaborations with new groups. So look for some new projects in the TWC.

This AGU was the first time that I’ve given a presentation at a conference of this size. I’ve given shorter talks at other, smaller gatherings, but nothing like this. It was a great experience. Great feedback from folks in the audience, good questions. I look forward to doing more presentations in the future. Thanks to James Michaelis for doing most of the writing for this presentation. I look forward to working on this project with him.

My big focus this time, more that I heard more talks and saw more posters on the topic, was data citation and attribution. Peter Fox posted to the TWC blog on this topic earlier. Data citation and attribution, as well as document citation and attribution and other forms of media and data, is something that people are interested in. As Peter mentioned, more from the standpoint that people want their work to be cited, to get credit for all the hours and resources that they put into their work. And I can certainly appreciate this. I’ve been looking at it also from the standpoint of data storage, persistence, and metadata associated with the data. We have even been chatting with folks from the RPI library about this, making sure that RPI resources are in a common place with common access methods, proper attribution and citation. Instead of each department, or each constellation, or each class, or each professor and student keeping track of their own work and storing them on their own machines, it would be nice to have a single location for the university where anyone can post their data, add semantic information about the data, and have that data searchable using varying means. Of course, it would be nice to utilize the semantic tools that are are creating at the lab.

So, I went to some of these sessions, found some of the posters, had some conversations with people, and brought back some business cards and scaled down versions of posters. I look forward to continuing our work with the library and with Jim Myers up on campus with this topic, and possibly bringing in other collaborators.

Follow

In the past few weeks, and during the conference, I collected a lot of opinions about the American Geophysical Union (AGU) Fall Meeting.  Some said it was their most important week of the year, others said it was unlike any other conference they attend, a conference on steroids if you will, and I think there was a general consensus that the event was downright exhausting.  I can’t comment with the second point, as this was the first conference that I’ve attended; however, I agree that this was easily the most important week to my research to date, and, yes, I am exhausted.

The week was exciting.  I got to meet a lot of people interested in eScience and informatics.  There was a broad coverage of research topics in the Earth and Space Science Informatics sessions, covering nearly every aspect of data management and curation.  There were presentations on the applications of data transfer formats, including netCDF and HDF, and conformance to metadata standards, such as ISO 19115.  Beyond simple data and metadata representations, many were interested in interoperability techniques, including OPeNDAP for application-level interoperability of data, and also vocabulary-level interoperability, which I found was a much more difficult concept for data curators to find value in.  Data curators are not yet seeing the importance of vocabulary interoperability because the “killer app” that utilizes integrated vocabularies is lagging far behind the initiatives towards interoperability.  I find this to be surprising as the “killer app” that many are looking for is a simple portal for spatio-temporal data search.  The major hindrances to the development of a spatio-temporal data search portal is not only metadata interoperability (through standards and vocabularies), but also search efficiency.  Technologies like Apache Solr are emerging for distributed indexing and query for keyword search over the metadata of deep web holdings, but spatio-temporal indices are being included in these technologies as an after thought.

Beyond the data/metadata/vocabulary research presented at the conference, there were numerous presentations on the importance of standardized web services, and many applications demonstrating the use of standards developed by the Open Geospatial Consortium, Amazon, and Open Archives Initiative, to name a few.  In the many presentations, I don’t think there was a clear indication of what the most popular web service standard was, nor was there an indication that the diverse set of available standards would converge.  Actually, Dan Pilone and his collaborators from NASA’s Earth Observing System (EOS) Clearinghouse (ECHO) put together a nice set of posters comparing the utility of web service standards from the perspective of both the developer and the end-user.  All of this work was particularly interesting to my research, which is starting to converge towards application- (or service-)level semantics.

My poster presentation was more inline with those focused on the applications and integration of web services.  I presented S2S, an application level framework for building customized user dashboards for data search and visualization that supports techniques such as advanced search and faceted browse.  The poster was well received, and I gathered both positive feedback and, to an extent, skepticism.  I spent the afternoon explaining the benefits to be reaped from application-level semantics, the successful application of the Semantic Web Methodology and Technology Development Process, and, for the skeptics, justifying the complexities of an application ontology.  I was able to compile the input of the conversations surrounding the poster, realizing the need to incorporate a data abstraction into the application ontology, and the need to provide further abstraction on the service concept to support a broader range of web services.  I think such abstractions will help to dissolve interest in web service standard convergence, and could serve as a foundation to a “killer app” that will spark further interest in vocabulary interoperability, and application-level semantics.

The week was exhausting.  I find that I need to work on my ability to absorb information, as even just days after the conference, while putting together this blog, I feel fuzzy on the details of the many conversations I partook in.  One of the problems I faced was that I am not great at taking notes and participating in a discussion concurrently.  In addition to this, I found it hard to get the chance to sit down and reflect on conversations had, both because some conversations would span an hour or more, and also because I found I was moving directly from one conversation to the next (too many things to do in too little time!).  I plan to work on my ability to jot down notes in the midst of discussions, and will be weary of transitioning to new conversations without taking a break (to reflect) in future conferences.  If any of these reflections on my experiences have taken your interest, feel free to shoot me an email (or find me in the lab, TWCers!).

Follow

Have attempted to elucidate at least part of my research agenda for the coming year — blogged at: http://blogs.nature.com/jhendler/2011/01/03/semantic-web-new-years-resolutions

-Jim Hendler

Follow