Tuesday, July 27, 2010

2010-07-27: NDIIPP Partners Meeting, IETF 78

On July 20-22, I was at the NDIIPP Partners Meeting in Arlington VA, along with Martin Klein and Michele Weigle. The Library of Congress has not yet uploaded a public summary of the meeting, but there were a number of interesting additions to previous NDIIPP Partners Meetings (edit: the meeting slides are now available). First, there were keynotes from both the Librarian of Congress, James Billington, as well as the Archivist of the United States, David Ferriero. There was also a ceremony to commemorate the charter members (which includes ODU CS) of the National Digital Stewardship Alliance (NDSA). I don't think the NDSA has a canonical web site yet, so the iPRES 2009 paper by Anderson, Gallinger & Potter is probably the best available description (edit: LC has announced a NDSA web site).

There was a theme of exploring the questions about "why we should care about digital preservation". The Library of Congress debuted this video, now available on their YouTube channel:



And I presented two sets of slides about Memento. One was in a break out session and focused on some of the details of http transactions as well as TimeBundles & TimeMaps, but the first presentation was in a plenary session in which I closed with an example of why digital preservation is important. For a recent conference submission, a reviewer asked:

Is (sic) there any statistics to show that many or a good number of Web users would like to get obsolete data or resources?

The answer we presented was that replaying the experience, as visualized through web resources, can be more compelling than a summary. The example concludes with an example about Hurricane Katrina.



The break out session slides are available too:



The day after NDIIPP, I was headed to Maastricht, The Netherlands to attend IETF 78 with Herbert Van de Sompel. We are working on an RFC for Memento, the first step of which is writing an Internet Draft which we hope to submit to the IETF soon. We learned a great deal about the ID/RFC process and met with several people who will help guide us through the process. Thanks to Mark Nottingham, we were even able to pitch Memento at the httpbis working group (see the agenda). Initial feedback was cautiously positive, but we were told by several people "I look forward to reading the Internet Draft".

And just for Johan's amusement, I had Herbert take the picture of me in Maastricht, next to giant french fries...

--Michael

Saturday, July 17, 2010

2010-07-17: Microsoft Research Faculty Summit 2010

On July 12-13 2010 I was at the Microsoft Research Faculty Summit 2010 in Redmond WA. The agenda was exciting and one of the few conferences that where I've had real difficulty in choosing which of the parallel session to attend.

The first keynote was about Kinect for Xbox 360. The demos were very impressive and I had no idea that motion capture was ready for the home market. Check out the trailer at the MS site.

The next session I attended was about the "Bing Dialog Model". I must confess that I'm unconvinced on how different Bing is from Google. Here's a side-by-side comparison of each search engine on the query "Michael Nelson":


They seem nearly identical to me: the tri-panel layout (controls on left, content in center, ads on right), the link layout/colors (blue title, black summary, green URI), interspersed images, tabs at the top, etc. The extended summary Bing gives you when you mouse over a link region is nice, and some of the details are different, but this is basically the same interface. They did demonstrate some summarize content (e.g., current and comparative status in sports box scores), but I can't help but feel that this is propagating anonymous (e.g., not bookmarkable) Web resources. The "Los Links" commercials are funny (episode 1 & episode 2), but I can't help but feeling they are addressing a problem I don't think I have. I also realize that given my knowledge of the tools, my search behavior and strategies have likely adapted (cf. Sapir-Whorf).

Next up was me during a lunchtime brown bag seminar chaired by Lee Dirks where I gave a well-attended talk about Memento:



I was a bit worried about how much detail to include, but I think I found the right level. There were a lot of good questions and discussion afterward, including a discussion with Alex Wade's summer intern, Kevin Lane, that extended through the next two sessions. The next session had a nice presentation from Tara McPherson about the journal "Vectors" and its follow-on "Scalar". The closing keynote was from Turing Award winner Chuck Thacker about FPGAs. Not being an architecture person the talk went over my head, but you don't pass up opportunities to catch a lecture from someone like Chuck.

The second day opened with a prestigious panel about transforming CS research. Rather than summarize it, I recommend you read it. Next was DemoFest, where you could play with Kinect. Over lunch I discussed possible Memento & Zoetrope integration with Eytan Adar. The next session featured an exciting presentation from Walter Alvarez about "Big History" and the ChronoZoom interface -- very exciting stuff. That session ended with a presentation about WorldWide Telescope, which was impressive but suffered for being scheduled after Walter. The closing keynote was about the making of Avatar (some of the concepts can be found in videos from Sequence Mag and Wired).

More notes can found with the hash tag "#facsumm".

This was an excellent conference and I was very happy to have the opportunity to speak about Memento. Thanks to Alex and Lee for the invitation.

--Michael

Thursday, July 15, 2010

2010-07-15: AMS Cloud Physics and Atmospheric Radiation 2010



I presented a poster at the 2010 13th Conference on Cloud Physics 13th Conference on Atmospheric Radiation in Portland Oregon, June 28 - July 2. This was my first atmospheric science meeting in the 2 years since taking off from NASA to attend full-time graduate studies at Old Dominion University. It was good to be back and catch up on old and new atmospheric sciences research being conducted by my colleages and others.

This conference takes place every 4 years. There were approximately 300 hundred scientists from around the world in attendance of which 60-70 were from NASA Langley. This was one of our important conferences to showcase our latest cloud and radiation results and products. Clouds and the Earth's Radiant Energy System (CERES) group were well represented. It seem like everyone at Langley who works on CERES were there. I saw many familiar faces and met several new CERES folks.

My paper was entitled Alternative Method for Data Fusion of NASA CERES and A-TRAIN datasets: An Evaluation of Triplestore. This paper examines a different method for placing geospatial scientific data into a Triplestore and ultimately directly on the World Wide Web where they can be indexed, discovered, and addressed. The mapping of geospatial data (or fusion) to a geo-location and Triplestore queries were presented at a high level.

I was disappointed that I did not plan a day trip to Mount St. Helens. My friend, who went with his family, said it was spectacular. I did however managed a late afternoon trip to Mt. Hood. The 1.5 hour drive was very scenic. I took this picture with my iPhone. It was my first trip in Oregon and it wont be my last.

-- Louis

Tuesday, July 6, 2010

2010-07-06: Travel Report for Hypertext and JCDL 2010


As mentioned earlier I had two papers accepted at HT and JCDL. In June it was time to travel to the conferences and represent the Old Dominion University colors.
HT 2010 took place in Toronto, Canada from June 13th-16th and was hosted by the University of Toronto. The acceptance rate of 37% was slightly higher than last year but the number of registered attendees seemed comparable.
I was glad to be able to give the very first presentation since it secured the probably greatest audience of the entire conference. My slides are available through Slideshare.
The paper itself titled "Is This a Good Title" can be obtained through the ACM Digital Library and its content was covered in my earlier post.
My personal highlight of the conference was the keynote by Andrew Dillon. He argued that research on Hypertext today is shaped too much by the Internet and its (inter-)linked nature. Representing his iSchool he made a point to support cross discipline research and suggested calling for papers in a topic independent track at HT 2011.
My fellow student Chuck also presented his paper and shared a detailed report of the conference in an earlier post.

Random facts:
- HT 2011 will be in Eindhoven, The Netherlands
- a cruise is not a good idea for a conference dinner
- the CN Tower is impressive and dining 1,151ft above the city is a great experience
- it feels good to have now two papers published at a conference where even Tim Berners-Lee's paper got rejected (if you don't know what I am talking about read this book)

Just one day after coming back from Toronto I left for Brisbane, Australia. JCDL 2010 took place from June 21st-25th in Surfers Paradise at the Gold Coast south of Brisbane. It was held in conjunction with ICADL which meant you had to chose between three parallel sessions. Presentations early on in a conference are usually good and the audience volume typically diminishes the longer the conference lasts. Hence I again was glad to see my presentation scheduled in the second session overall.
I mentioned the paper (available here) in a previous blog and my slides are also available via Slideshare.
My two personal favorites were again keynote speeches. The first given by Katy Börner from Indiana University. Listing her titles and appointments would take too long (see her web site) but amongst others she is the founder of the Cyberinfrastructure for Network Science Center at IU. She gave an insight into a few of her projects related to data analysis and visualization. It is amazing how information access changes through visualization techniques.
David Rosenthal is the chief scientist of the LOCKSS project. In his keynote he argued for the need of new models and approaches of preservation focused on the dynamic and service aspects of the Internet. In his opinion the "old copy-and-re-publish model" works for collecting and preserving static content but not for the web of services. Publication and preservation need to be re-defined to keep up with the transition towards services. The entire keynote is available here.

Random facts:
- JCDL 2011 will be in Ottawa, Canada
- the Australian Outback Spectacular is neither spectacular nor at all suitable for a conference dinner
- US Airways has a hard time distinguishing between the airports in Sydney (SYD) and Syracuse, NY (SYR) and may send your luggage to either one
- the Gold Coast winter is very pleasant and "Surfers Paradise" is not exaggerated
--
martin

Monday, July 5, 2010

2010-07-05: Foo Camp 2010

I attended the 2010 Foo Camp in Sebastopol CA, June 25-27. For those who are unfamiliar, Foo Camp is an invitation-only "unconference" -- which is basically a conference that consists entirely of birds-of-a-feather sessions as well as the impromptu hallway and dinner conversations that make conferences useful.

There were approximately 250 people there and by my estimation they were mostly young (25-35) entrepreneurs (current and former). There were a smattering of others as well: artists, writers, professors, VCs, etc. The best way I can describe Foo Camp is a combination of Burning Man (culture of participation), SIGGRAPH (culture of demonstration), and a country club (culture of capitalism). Geeks aren't really known for being extroverted, but the format of Foo Camp pretty much requires meeting new people and interaction with people outside of your existing circle of colleagues. I was surprised at how approachable most people were.

Formulating the schedule consists of a big scrum on the opening night to place stickers on the schedule board. On Saturday the contents were transcribed to the wiki, but the actual schedule changed frequently. Sessions covered many topics and varied greatly in the number of attendees. The highlights of Friday and Saturday were the Ignite presentations: 5 minutes, 20 slides on auto-advance (15 sec each). The videos from Foo Camp haven't been uploaded yet to the YouTube channel, but hopefully soon.

Getting an invitation to Foo Camp is a pretty big deal, and you're reminded several times throughout the weekend that a return invitation depends on the quality of your contribution to Foo Camp. I was there to talk about Memento, which was the impetus for the Foo Camp invitation. Given the entrepreneurial focus of Foo Camp, the interest in archiving and preservation was obviously not as a strong as it is in most of the conferences and workshops I attend. Still, I made a few really good contacts that I will detail in the future.

I had a great time, but it was very exhausting. Johan, who has been to a couple of SciFoos, told me that few people camp. That may be true at SciFoo, but there were many campers at Foo Camp. Sebastopol is a pretty small town and the hotels were booked, so I drove in from Santa Rosa each day. I took a few photos but there are a handful of other photos and blog posts that are better capture the spirit: Laughing Squid, Dean Putney, Scott Berkun. The hash tag was "foo10".

Very special thanks to Tim O'Reilly, Sara Winge and the rest of the folks at O'Reilly for the invitation and event. Three days sounded like a lot at first, but it was over in a flash.

--Michael