
A couple pints at the Dog and Duck after the JCDL workshop. Excellent. #jcdl2009
#jcdl2009 is over now. lots of people going to see the austin bats tonight. should be interesting!
@HCIR_GeneG he is building a cival war archive to facilitate student generated content creation at a medium sized public univ. #jcdl2009
Today I learned about creating OS X Finder workflows - thanks to @dchud and #jcdl2009. A nice way to end the week.
Part of our workshop - build a librarian. Ours was tatted rebel librarian #jcdl2009
I should run this analysis tomorrow; as of 12:48 pdt, JCDL2009 tweet counts: 124 @HCIR_GeneG, 56 @gingdottwit, 41 @dchud, 29 @Alan_Wolf
From #jcdl2009 to #openvideo, Wikipedia is always on the schedule
Morrill: which small segment of your userbase do you wanna 'own'. and 'own' them. #jcdl2009
@gingdottwit @alan_wolf nice to follow along w/ workshop discussion after leaving #jcdl2009
Morrill: pizza hut starved students of pizza for 2 months, and then asked them what type of pizza they wanted. funny. #jcdl2009.
Morrill - developing a target user begins by segmenting the population, segments are stable, identifiable, measurable, predictive #jcdl2009
Morrill - example - the idea store in the UK - focused on younger, tech interested users, but brought in others in the community #jcdl2009
morrill - focusing on target users does not need to exclude other users. Orgs that focus on a target user get many other users #jcdl2009
Morrill: if you try to support everybody then you struggle to properly support anyone. focusing on one group wont exclude others #jcdl2009
Morrill if your target user is everybody, there is a good chance that your actual users will be nobody #jcdl2009
Morrill - selecting a target user deciding whether to be all things to all people or specialize #jcdl2009
Morrill is back again, what an over-achiever ;-) If you build it, they will come...maybe... #jcdl2009
george: its not just students: if you have keen study participants. recruit them to support the evaluation too! #jcdl2009
Sally Jo Cunningham - collaborative music selection at social gatherings- 30 student observers generated 250 pages of observations #jcdl2009
Sally Jo Cunningham - Distribute data collection - uses students as interviewers #jcdl2009
sjc: 80 students x 2 interviews. (for example) is LOADS of interviews. quality is often questioned. but diversity of Qs is rife! #jcdl2009.
Sally Jo Cunningham - lightweight method - Mine existing data sets, example music playlist construction #jcdl2009
sjc: method 3: share the load. have lots of students? get them to get data. its good practice for them! and you. #jcdl2009.
sjc: in previous experience: interviewees tell you about their 'serious' mixes of music. artofthemix.org gave fun examples, etc. #jcdl2009
sjc: method 2: mine existing data, rather than generating new data. wanna know about music collections? hit artofthemix.org. etc. #jcdl2009
small studies w/multiple mthds to get big amts of data example from Rochester http://bit.ly/bAv2e - Sally Jo Cunningham #jcdl2009
sjc: method 1: lots of small studies rather than one large study. break down your big question into multiple aspects. #jcdl2009
sally jo cunningham on 3 ways to get maximal usability data from minimal effort #jcdl2009. nice.
morgan and @Alan_Wolf - actually tagclounds or wordles make great coding guides for survey responses and interviews. #jcdl2009
morgan and @Alan_Wolf have been self studying their surveys for size/validity/etc. Refine/reduce your needs+questions unfront #jcdl2009
morgan and @Alan_Wolf talking to us about refining/optimizing surveys to make them as lightweight as possible #jcdl2009
FYI - @gingdottwit and @Alan_Wolf are tweeting about the lightweight user evaulation methods workshop at #jcdl2009.
Amanda Xu integrating surveys into research #JCDL2009 multiple audiences in evaluation design and their perspectives #jcdl2009
Amanda Xu: talking about how planners, owners, designers, builders, integrators, and users, each view 6 eval criteria differently. #jcdl2009
Need to consider the issues of rewards and participation in research participants - need to aggregate research #jcdl2009
Morrill and McMartin talking about running user panels to evaluate DL interfaces #jcdl2009
National Science Foundation evaluation guide - http://bit.ly/sxbr1
RT @Alan_Wolf: Logic Models Also guides from NSF and CDC. http://bit.ly/1xHFmz - Kellogg Foundation evaluation toolkit #jcdl2009
Flora Mc Martin is talking about Logic Models http://bit.ly/1xHFmz - Kellogg Foundation evaluation toolkit #jcdl2009
Posted "Tweeting at JCDL" http://palblog.fxpal.com/?p=1166
brain is equivalent to clotted cream after 3 days of JCDL
just watched a introduction to SEASR in the Digital Content workshop at #jcdl2009 and is amazed. Trying to resist the urge to scrap my code.
Max Wilson - very interesting tool for evaluating user interface of search interfaces, well worth a look. http://mspace.fm/sii/ #jcdl2009
In the Second Generation Digital Library workshop at JCDL -- looks like
Max Wilson - Evaluating search user interfaces - Sii - search interface inspector http://mspace.fm/sii/ #jcdl2009
George Buchanan - Avoid casting too wide a net in evaluation - Yes - a too much information problem. #jcdl2009
Buchanan, lightweight studies mean not thinking broad/vague, but about specific/targeted use. More on this later in the workshop.&#160;#jcdl2009
George Buchanan - lightweight evaluation as evaluation with a lot of scaffolding # jcdl2009
interesting crowd at the Lightweight User Eval workshop at #jcdl2009. Should be good.
Mick Khoo is starting the lightweight evaluation for digital libraries workshop #jcdl2009 it'll be interesting to see how it folds together
getting ready for a workshop on lightweight evaluation methods for digital libraries at JCDL in Austin, Tx
http://twitpic.com/7reuy bbq at Rudy's brisket, sausage, ribs http://bit.ly/J117U. #jcdl2009
RT @dchud: Lesk: "nobody from Amazon or Google is here, so this conference is irrelevant" #jcdl2009 - I don't think world is driven by 2 Cos
Is glad that the last conference day at #jcdl2009 is done. Any more conference and I would have had more things to do than time in a day.
Wait, another last tweet: What was the award winning paper? #jcdl2009
JCDL 2010 is in Queensland, hmmm. Have to see if I can swing that. #jcdl2010 papers due jan 31 2010
In some way this last session should have been the first one. The discussion has just started. Frustrated. #jcdl2009
Lest - issue of small number of companies in control. Already a problem, if your amazon acct cancelled, lose some kindle function #jcdl2009
Cliff Lynch: the notion that something could be 'unpublished" is problematic in the US. #jcdlgoogle #jcdl2009
We should worry about becoming dependent on only a few information resources and services. #jcdl2009
Michael Lesk: we are becoming dependent on only a few companies. Sensorship is dangerous. #jcdlgoogle #jcdl2009
We're almost out of time and finally the discussion starts to open up. #jcdl2009
Gretchen Hoffman: targeted ads will require data collection. This scares people. #jcdlgoogle #jcdl2009
Cliff Lynch: given amount of time people spend watching or ignoring ads, ads are preserved incredibly poorly. #jcdlgoogle #jcdl2009
@dchud I have the same kind of feeling. #jcdl2009
Cliff Lynch: advertising as erosion of attention. People tune most of it out. #jcdlgoogle #jcdl2009
@ccmarshall Advertising is linked to future of publishing. companies pay attention to ad placement algos! #jcdlgoogle #jcdl2009
this sort of open-ended panel discussion always leaves me feeling cold and angry. this one's not an exception. #jcdl2009
Cliff Lynch: fragmentation inevitable, mass culture a function of limits of technology. #jcdlgoogle #jcdl2009
RT @taswegian: Provenance, proof and trust - Upper layers of the SW, #semtech2009 [#jcdl2009 : future place for DLib conversation?]
Gretchen Hoffman: easier to avoid uncomfortable info online. #jcdlgoogle #jcdl2009
@HCIR_GeneG Google came out of NSF's DL funding in the 90s, but Lesk is right, DL is increasing irrelevant due to lack of funding.#jcdl2009
Michael Lesk: should look at how to run groups without vandals taking over.(e.g., Wiikipedia) #jcdlgoogle #jcdl2009
Q: is intellectual fragmentation of society inevitable? #jcdlgoogle #jcdl2009
Michael Lesk: what papers presented here matter to Amazon &amp; Google? DL is irrelevant. #jcdlgoogle #jcdl2009
Gretchen Hoffman: Google's settlement makes this difficult. #jcdlgoogle #jcdl2009
Gretchen Hoffman: if author doesn't step forward, author doesn't care. should go into public domain. #jcdlgoogle #jcdl2009
Cliff Lynch: need holistic approach that deals with all cultural materials (games, books, images, etc.) #jcdlgoogle #jcdl2009
Gretchen Hoffman: Law is slow, intended to be flexible. Need to battle it out, not satisfying. #jcdlgoogle #jcdl2009
Issue of orphaned works in game world is huge #jcdl2009
Q: too many different contracts may confuse users. #jcdlgoogle #jcdl2009
Ben Bederson: Fair use works well for Kindle: can download a free sample of any book. #jcdlgoogle #jcdl2009
Gretchen Hoffman: not right that google should have to fight for the right to do the kind of scanning it is doing. #jcdlgoogle #jcdl2009
At the Google as library redux JCDL ending panel
Gretchen Hoffman: such attitudes force publishers to react, to find other ways to make money. #jcdlgoogle #jcdl2009
Gretchen Hoffman: freshmen don't get significance of intellectual property. #jcdlgoogle #jcdl2009
Michael Lesk: why have a science library if all disciplines do their own archival? #jcdlgoogle #jcdl2009
Michael Lesk: disappointed that library communicty doesn't handle scientific databases. #jcdlgoogle #jcdl2009
Cliff Lynch: can online advertising be more effective if done properly? how to avoid flash-based garbage? #jcdlgoogle #jcdl2009
Cliff Lynch: print ad rates are much higher than online rates. Why is that? #jcdlgoogle #jcdl2009
Michael Lesk: need experiments to show new ways of doing things. #jcdlgoogle #jcdl2009
Ed Fox: would like to have books available to the blind in audio to be available to others as well. #jcdlgoogle #jcdl2009
Read @Charlie_Haas 's novel "The Enthusiast" about fragmenting cultural interests and coherent society for fascinating POV #jcdl2009
Cliff lynch: small groups can obtain critical mass; big gap between that and coherent society. #jcdlgoogle #jcdl2009
Cliff Lynch: how fragmented can news get before there is general distrust of the source? Reinforcement important. #jcdlgoogle #jcdl2009
Cliff Lynch: audience size for any program is much less now than in the 60s -- many more channels. #jcdlgoogle #jcdl2009
Cliff Lynch: facts suggest much of mass audience is already lost. #jcdlgoogle #jcdl2009
Gerhard Fischer: important questions or what you read, not how you read it. #jcdlgoogle #jcdl2009
Regrets my question about trust wrt Google and the Japanese maps locating the burakumin is now irrelevant in the panel discussion #JCDL2009
Gerhard Fischer: does personalized newspaper destroy common understanding, sharing experience? #jcdlgoogle #jcdl2009
Gerhard Fischer: is the medium (book/device) an important aspect of reading? Is human attention the limiting resource? #jcdlgoogle #jcdl2009
Cliff Lynch: secondary markets for books may be wiped out; nasty unanticipated effects. #jcdlgoogle #jcdl2009
Michael Lesk: Euopean parlaimant now has Pirate Party dedicated to repeal of copyright law #jcdlgoogle #jcdl2009
Gretchen Hoffman: How will this affect the law? Authors can be compensated, but not through current model. #jcdlgoogle #jcdl2009
Gretchen Hoffman: business model for books may change to ad model. #jcdlgoogle #jcdl2009
Gretchen Hoffman: copyright law enables contract law. Law needs to be modified. #jcdlgoogle #jcdl2009
q: is copyright irrelevant vs. contract law? #jcdlgoogle #jcdl2009
q: publishers fear loss of control over IP #jcdlgoogle #jcdl2009
q: publishers can choose to enable audio. National Library Service can record audio without royalties. #jcdlgoogle #jcdl2009
q: legal implications of convergence of devices? Kinbdle and iPhone don't work well together. #jcdlgoogle #jcdl2009
Gretchen Hoffman: law is a gigantic club, not balanced, probelm for achieving our goals. #jcdlgoogle #jcdl2009
Cliff Lynch: specific accommodations have been made in copyright law to make books available to the blind. #jcdlgoogle #jcdl2009
Cliff Lynch: that's why authors started panicking -- separate stream of revenue #jcdlgoogle #jcdl2009
NEMA meeting at JCDL. Music informational retrieval community round tripping from XML to RDF and back
Cliff Lynch: speech synthesis is much worse than recorded audio, but getting better. Soon should have plausible audio. #jcdlgoogle #jcdl2009
Michael Lesk: Amazon retracted voice from books based on authors' complaints. #jcdlgoogle #jcdl2009
q: blind people prefer spoken to braille books. Implications? #jcdlgoogle #jcdl2009
Cliff Lynch: searching has replaced piling #jcdlgoogle #jcdl2009
Michael Lesk: journals have moved to digital, but they are printed for careful reading. Hoarding of paper is gone. #jcdlgoogle #jcdl2009
Gretchen Hoffman: we do need to find a new way of thinking about the space, including all the players. #jcdlgoogle #jcdl2009
Q: not book vs. paper dichotomy; both important #jcdlgoogle #jcdl2009
Mike Lesk says that google books (and JCDL) are irrelevant. Amazon digital books is where the action is.
Cliff Lynch: What do maintainers of collections need to do to facilitate integration into future DL environemnts? #jcdlgoogle #jcdl2009
Cliff Lynch: integration of fragments of cultural materials from different traditions is a challenge #jcdlgoogle #jcdl2009
Cliff Lynch: How soon and how much will reading on screen happen? More skeptical than Lesk about vanishing of book. #jcdlgoogle #jcdl2009
Cliff Lynch: subversion of authorial voice vs. contribution to knowledge bases. #jcdlgoogle #jcdl2009
Cliff Lynch: scientific data suggest a different future. much more interlinked, active, multimedia. #jcdlgoogle #jcdl2009
Cliff Lynch: will libraries be PDF+search, or will they be knowledge environments? #jcdlgoogle #jcdl2009
Google settlement is adopted only mitigates the problem of orphaned works in books (in the USA), not other media Cliff Lynch #jcdl2009
Cliff Lynch: settlement won't fix orphan works problem; limited to books. Photos, art, sound, not convered #jcdlgoogle #jcdl2009
Cliff Lynch: fair use was not trial of century, settlement puts the issues aside. #jcdlgoogle #jcdl2009
Cliff Lynch: Expectations of Google Book Scanning project were not met #jcdlgoogle #jcdl2009
Future is skimming, searching and snippets, multimedia, and long tail - michael lesk #jcdl2009
Lesk: "nobody from Amazon or Google is here, so this conference is irrelevant" #jcdl2009
Michael Lesk: nobody here from Amazon or Google. This conference this irrelevant! #jcdlgoogle #jcdl2009
Michael Lesk: Future is skimming, snippets, not reading. #jcdlgoogle #jcdl2009
Michael Lesk: study compared paper book, pdf book: a lot more pages read in PDF #jcdlgoogle #jcdl2009
Michael Lesk: Will Google have to license to Amazon? #jcdlgoogle #jcdl2009
Michael Lesk: It's about Amazon, not Google. 3B books bought, 2B borrowed from libraries. 40Ml from research libraries.#jcdlgoogle #jcdl2009
Michael Lesk: books are not really paper #jcdlgoogle #jcdl2009
Michael Lesk: Do readers want paper or pixels? #jcdlgoogle #jcdl2009
Gretchen Hoffman: publishers, congress don't necessarily get the new situation. #jcdlgoogle #jcdl2009
Gretchen Hoffman: search engine is fair use #jcdlgoogle #jcdl2009
Gretchen Hoffman: user expectations, behavior, technology industry, content industry, regulation all affect each other #jcdlgoogle #jcdl2009
JCDL Panel on Google as Library: Getchen Hoffman, Clifford Lynch, and Michael Lesk #jcdl2009 #jcdlgoogle
speaker just bagged a digital object live, onstage. oh, the humanity! #jcdl2009
several papers on Warrick and reconstructing web sites: http://www.cs.odu.edu/~fmccown/research/lazy/ #jcdl2009
Two (very raw, very basic) LDA topic analysis (using MALLET) visualizations for #jcdl2009 papers: http://bit.ly/b3F51 http://bit.ly/ItZfE
Flashing back to my Internet Archive days, wondering if there are IP address assignment archives anywhere. Or is all that lost. #jcdl2009
talk: EverLast, adding temporal query to web archives http://www.mpi-inf.mpg.de/~aanand/jcdl2009.pdf from Max Planck Inst. #jcdl2009
Konishiki! at #jcdl2009! or, at least, on the screen.
grateful for J McDonough's talk "aligning METS w/OAI-ORE" pointing out issues but it seems they serve diff purposes, so why align? #jcdl2009
at JCDL learned that Austin has a company called OAI!! OAI
We're here at Day 3 of JCDL. First up: ALigning METS with the OAI-ORE Data Model
is wondering if he is going to make it to the morning session at #jcdl2009 at this point.
right. i know its morning back home. but im finally going to bed before an exciting session tomorrow morning at #jcdl2009
JCDL 2009: Session 6 http://bit.ly/xNH9l Cheers Judith!
Schneider, Jodi: JCDL 2009 Poster Session in Second Life http://snipurl.com/kck1a
@_cb_ #jcdl2009 SL poster session screenshots http://jodischneider.com/blog/2009/06/18/jcdl-2009-poster-session-in-second-life/ @joannascott
RT @MSFTResearch Answering the question *What we should Preserve from a Digital World*, MS researcher Cathy Marshall at JCDL http://www....
@christophburke I am at JCDL2009 (Joint Conference for Digital Libraries, it's an International thing)
Thinking about going to see the bats #jcdl2009
@jcdl2009
Patricio Galeas: # of coeffficients determines passage size for computing similarity #jcdl2009
Answering the question &#8220;What we should Preserve from a Digital World&#8221;, MS researcher Cathy Marshall at JCDL http://www.jcdl2009.org/ in TX
Miles Efron: in effect, this a dimensionality reduction approach #jcdl2009
Patricio Galeas:q: why use FT? a: FT compression due to only some coefficient, also defines neighborhoods #jcdl2009
Patricio Galeas: neighborhood models outperforms other models using 40 terms and top-10 documents #jcdl2009
Patricio Galeas: query expansion based on term position: use neighborhood of terms to expand queries #jcdl2009
anybody at #jcdl2009, who looks at twitter, a practitioner working regularly on designing DL interfaces?
Patricio Galeas: ranking optimization max similarity between query and function that describes the document. #jcdl2009
Patricio Galeas: low-order Fourier coefficients approximate term positions. Then use dot product to estimate similarity. #jcdl2009
Patricio Galeas: Positions are summation of delta fcns over document length. Replace all instances iwth a Fourier decompositon. #jcdl2009
Patricio Galeas: use position of terms in matching documents to estimate relevance #jcdl2009
hit the MAX_INPUT wall after the facebook talk made me look at facebook. actually leaning against the wall now in the far back. #jcdl2009
Patricio Galeas: frequency and distribution are predictors of relevance #jcdl2009
Patricio Galeas: Document Relevance Assessment via Term Distrubtion Analysis Using Fourier Series Expansion #jcdl2009
The Perseus project fragmentary author, fragmentary history talk was a speedy whole. #JCDL2009
abdi asks if difficulty of call-centre problem should have a weighting effect on the model. answer - yes probably. #jcdl2009
Mingfang Wu: both kinds were seen #jcdl2009
Mingfang Wu: Q: does type of query (known-item vs. exploratory) benefit differently from this approach? #jcdl2009
@HCIR_GeneG asks if mentees learn from mentors, yes mentors are suggesting better queries to people, aswell as re-ranking results. #jcdl2009
Considering staying until Friday and using the extra time to explore Austin by car a bit, but most of my group is leaving early... #jcdl
Mingfang Wu: queries that don't have relevant documents do not benefit from approach #jcdl2009
Wu - simple answer. if a query is repeated 9 times or more, then the cost/benefit (in wage/time) is worth it. #jcdl2009
Mingfang Wu: Model suggests if query repeats at least nine times, it deserves attention from mentor #jcdl2009
Wu's call center mentor scenario. the cost of reranking is occured by mentor, and the benefit is for the supervisee. hard scenario #jcdl2009
Mingfang Wu: use time x hourly rate of different roles #jcdl2009
Mingfang Wu: cost=time spent to make assessment; benefit=time saved reading irrelevant documents. #jcdl2009
Mingfang Wu: reranked based on mentors' judgments of relevance; improved precision of 144%, 90$ and 41% at 1,3,10 documents #jcdl2009
Mingfang Wu: observed high agreement (78%) among mentors #jcdl2009
Mingfang Wu: Study: mentors judged relevance of randonly-ordered documents returned by 50 queries drawn from the query log. #jcdl2009
Mingfang Wu: But this approach has significant personnel costs. How to decide when to intervene? #jcdl2009
Mingfang Wu: mentors can spend time to re-order the list of results forwarded to the first-line call handler. #jcdl2009
Mingfang Wu: Example of collaborative information seeking between call handlers and more senior consultants (mentors) #jcdl2009
@HCIR_GeneG always seems to be seconds ahead of me when tweeting about #jcdl2009
Mingfang Wu is talking about allowing users to manually change result-list-ordering in enterprise search systems. #jcdl2009
Mingfang Wu: What is the cost of human intervention to improve quality of results in a call center? #jcdl2009
Mingfang Wu: Cost and benefit Analysis of Mediated Enterprise Search #jcdl2009
@plbogen i don't know about you but i resemble that remark :) #jcdl2009
Frank McCown at JCDL talks about how to scrape your Facebook data - because ... how long will it last?
Oh, did anyone else get the venn diagram from the "After Facebook " talk ? #jcdl2009
was just told that twitter is the intersection of stalking, adhd, and narcissism by Frank McCown at #jcdl2009.
really enjoyed another Cathy Marshall talk at #jcdl2009, as usual.
many people in the audience seem to be not convinced by @ccmarshall at #jcdl2009.
@gingdottwit re: tag with own tags; search with controlled terms. Strange, don't existing systems encourage the opposite? #jcdl2009
For the record, Cathy Marshall is @ccmarshall #jcdl2009
marshall says folksonomies are NOT as useful as people THINK. #jcdl2009
marshall says folksonomies are NOT as useful as people thing. #jcdl2009
Cathy Marshall: tags don't have the folksonomc power that people say they do #jcdl2009
Cathy Marshall: location tags were not localized enough (Europe, italy, rather than Milan, Galeria); geotagging might be better #jcdl2009
i swear, if Cathy Marshall were giving this #jcdl2009 talk at #code4lib all present would be exploding w/giggles
Cathy Marshall: tags for personal content, titles summarized, captions told story from guide book #jcdl2009
marshall says tags tell you about the place and artefacts. photo titles and descriptions tell you about the contexts and stories. #jcdl2009.
Cathy Marshall: titles better than tags for finding picture #jcdl2009
Cathy Marshall: tags weren't as similar as the titles! #jcdl2009
Cathy Marshall: flickr has many instances of the same picture; might be useful to analyze tags on those pictures #jcdl2009
neat hack: Cathy Marshall studied reader annotations in many copies of same used course textbooks #jcdl2009
marhsall wants to put the question 'do folksonomies work?' to rest. lets put it behind us. #jcdl2009
Cathy Marshall: Want to answer the loaded question "Do folksonomies work?" #jcdl2009
Cathy Marshall's best paper nominee paper on the use of tags on flickr images. #jcdl2009
marshall begins on comparing user tags to other user-generated metadata. #jcdl2009.
these last2, and the next paper, are best paper finalists #jcdl2009. auto-assessment of quality. OCR triangulation, and c.marshall on tags.
@gingdottwit I noticed that people weren't identifying themselves. session chair should prompt that. #jcdl2009
Will Lund is talking about triangulating multiple OCR methods for better auto-archiving of old docs. #jcdl2009
#jcdl2009 talking about problems with ocr...i am curious how prime recognition ocr sw deals with guessing
I seem to be the only1 preluding Qs with my name/affil. is that not normal at&#160;#jcdl2009 like it is elsewhere?
Jun 16, 2009 IEEE/ACM 9th Joint Conference on Digital Libraries (JCDL).. http://garysguide.org/2360292
Bathard et al got 2 novice quality assessors to study 1k resources and found 5 key indicators/differentiators. used for autocheck. #jcdl2009
Just published JCDL 2009 ( http://bit.ly/56fO0 ) #thenoisychannel
Bathard identified, using coding analysis, 25 dimension relating to quality. then found 12 key ones. appropriate graphics is top. #jcdl2009
ZATAMM meets JCDL. Hmm. #jcdl2009
#jcdl2009 automating quality evall for dl collections
Bathard will go on to talk about a machine learning model based on the way experts assess resources. #jcdl2009
Bathard is talking about how you can scalably assess the quality of resources submittd to educational resources. studying experts. #jcdl2009
Enjoying #jcdl2009 Tweets. This is one of the best uses of Twitter (Tweeting live event for those who can't be there).
@HCIR_GeneG I'll post a list of links to #jcdl2009 papers on #thenoisychannel when there's a critical mass of them on authors' web sites.
study suggests users prefer to tag w. own tags, but search via controlld vocab. is web2.0 librarian role now curation tags-&gt;vocab? #jcdl2009
Timely Essay by Jeremy John "The future of saving our past" Nature 459, 775-6 11 June 2009 doi:10.1038/459775a #jcdl2009
@edsu she said most of it is restricted access but i forget how/why, would be great if that were less so :( neat project, though! #jcdl2009
a million chinese books project - 100s of TBs worth http://www.cadal.zju.edu.cn/Index.action #jcdl2009
@band thanks! graphviz is an endless supply of eye candy. this might read better: http://www.flickr.com/photos/dchud/3635453183/ #jcdl2009
graphs of mallet topics from #jcdl2009 papers http://www.flickr.com/photos/dchud/3636247610/ http://www.flickr.com/photos/dchud/3635432831/
@dtunkelang have you read Clarkson et al "Generalized Formal Models" paper from jcdl? looks interesting but it's not online.
Given the motivation of GeM, will the Impatient User wait for the statistical analysis required for the fusion? #jcdl2009
@_cb_ cool! am drawing a graph. will do one for yours, too. #jcdl2009
@dchud here's my stab at mallet topics: http://paste.lisp.org/display/81998 -- might want to strip boilerplate, orgs, etc #jcdl2009
@judielaine 203 - talk on chinese character analysis is looking good :) #jcdl2009
Can't decide between sessions. Will croud source the choice, I guess. #jcdl2009
Gerhard Fischer: "mindless trust is more prevalent in experts-based cultures"... THE nugget of wisdom during this loooooong talk #jcdl2009
mallet's output, keys from #jcdl2009 papers: http://paste.lisp.org/display/81957#1
Side effect of participatory culture: contrast between expertise and punditry. #jcdl2009
We're at JCDL 2009! Day 2 Plenary wrapping up.
Gerhard Fischer: one of basic skills to be taught is how to recognize trusted and reliable sources. #jcdl2009
How to teach how to recognize trusted and reliable resources? #jcdl2009
just used mallet to extract topic model from #jcdl2009 papers. i think. any other attendee willing/able to talk me through how this works?
Gerhard Fischer: how can we establish trust and reliability in participatory cultures? #jcdl2009
Gerhard Fischer: mindless trust in experts is a problem #jcdl2009
is trying to set up Fedora Commons while waiting for Session 7 to start at #jcdl2009.
Fischer: differences between standardization and improvisation. #jcdl 2009
@hcir_geneg tweeting and listening is a problem. #jcdl2009
Challenge: how to tweet and listen at the same time? #jcdl2009
Gerhard Fischer: how to combine head (basic knowledge and skills) with tail (personally meaningful problems) in learning. #jcdl2009
Gerhard Fischer: long tail in learning is personally-meaningful problems #jcdl2009
@dtunkelang #jcdl2009 continues to be very interesting. theres an enterprise search paper later today, i think.
Gerhard Fischer: democratic culture: "publish and filter" encourages more participation, but has issues of trust and reliability. #jcdl2009
Gerhard Fischer: consumer culture: problem with "filter and publish" is making all voices heard. #jcdl2009
Gerhard Fischer: 21st century problems are general systems and multi-disciplinary problems. #jcdl2009
hearing lots of talks about SVM/LDA at #jcdl2009, see http://mallet.cs.umass.edu/ and http://mallet.cs.umass.edu/ (via @_cb_, @benosteen)
Gerhard Fischer - moving from social productivity to social creativity #jcdl2009
My reaction to Fischer's claims: Participation would be encouraged more by task-tailored rather than generic tools. #jcdl2009
But my take is that SketchUp is so difficult to learn that in some ways it discourages participation. #jcdl2009
I'm jealous of the folks attending JCDL http://www.jcdl2009.org/
Gerhard Fischer uses SketchUp as an example of tools that enable culture of participation #jcdl2009
link to just-mentioned D-Lib article: http://www.dlib.org/dlib/may02/wright/05wright.html #jcdl2009
Gerhard Fischer talking further on the emphasis for public participation and encouraging ownership of information problems. #jcdl2009
Gerhard Fischer suggest 'infoware', from ischools, will be more important for digital libraries than software inc HCI focus #jcdl2009
fischer's culture of participation: evolution of digital libraries : hardware -&gt; software -&gt; infoware; EE -&gt; CS -&gt; SI #jcdl2009
@judell do you know about these folks? http://www.language-archives.org/ met some of them yesterday at #jcdl2009
Another wait at the airport. On my way to Austin for #jcdl2009 workshop and some meetings.
@plbogen yes the posters are still up at the Digital Preserve on Second Life #jcdl2009
Lots of other work showing different users have different needs (seniors, novice scholars, unsuccessful queriers, etc) #jcdl2009
Highlights from #jcdl2009 poster session: CMS Informedia, UT University Collections, MetRe (metadata revision process), LCSH/wikipedia xwalk
Just finished presenting the poster for JCDL
@mantruc Is the virtual #jcdl2009 poster session still up?
I'd say our virtual poster session went quite well, thanks to all who came in! #jcdl2009
@_cb_ yeah I took screenshots and will blog #jcdl2009 poster session soon. Great conversations, interesting visuals, glad I went.
interesting poster session some great posters for all interests. I liked the study of how users deal with 0 results in pubmed #jcdl2009
in Second Life for a poster session #JCDL http://tiny.cc/43Zws
Standing at my jcdl poster in second life sans wine and snacks
@pmedward dont miss our Second Life demo at #jcdl2009 going on live now!
@dchud *this* is the list i expected when extracting keywords from this conf's papers' texts: http://paste.lisp.org/display/81957 #jcdl2009
Visiting SecondLife for the #jcdl2009 poster session http://bit.ly/rJxqA
Shameless plug: stop by poster 190, back-right corner of the room. #jcdl2009
Curious.Judith &#187; Blog Archive &#187; JCDL 2009: Panel 1. What should we ... http://bit.ly/ZplBU
elvis in minute madness at #jcdl funny.
Elvis has arrived at #jcdl2009
#jcdl2009 Minute Madness about Demos and Posters ... in three ... two ... one ... GO!
The JCDL poster session begins at 5:45 CDT http://bit.ly/rJxqA
TDL: lots of work with OAI-PMH + ORE for metadata + content replication, flexible (+scalable) architecture #jcdl2009
Texas Dig. Lib has clearly put a lot of thought (and resources) into their ETD workflow/management/infrastructure #jcdl2009
thanks for the outloud citation ed clarkson. good talk. im gonna need to read the paper to understand your model tho :) #jcdl2009
facets: looks like an interesting paper; relational calculus is clearly above my head #jcdl2009
facets structural design: hierarchy, indirect facets (facets about facets), single vs multi focus #jcdl2009
talk on faceted navigation at #jcdl2009. right up my street. Go Edward Clarkson.
facets interaction design considerations: cascading + cardinality; #jcdl2009
formal models for faceted UI; faceted navigation survey, features, best practices to inform software development #jcdl2009
Writing while attending the #JCDL2009 sessions is working : http://www.grey-cat.com/curious has notes up.
@_cb_ neat talk, thanks for posting the video! #jcdl2009
video cmodel pres done; mp4 slides: http://bit.ly/dxIFO (launch from QT player) + pdf: http://bit.ly/9N2gg (less interesting...) #jcdl2009
@plbogen it would be good to surface the other associations being raised here as well. #jcdl2009
loving the idea of music info retrieval by chord sequence alignment #jcdl2009
keeps being reminded of old Neal Stephenson novels by the panel discussion at #jcdl2009.
Answer to what we'll preserve in a born-digital world: everything. Except for the stuff we loose (which might be a relief) #JCDL2009
Glad that A.G. mentioned power [politics? control?] influencing what gets collected and preserved. #jcdl2009
wondering, if the grouping UI of ViGOR reduced number of queries needed, is this going to be useful for mobile technology? #jcdl2009
ViGOR: ran a test with TRECVID http://bit.ly/11R1oA ; interesting video corpus for discovery, users still preferred ViGOR #jcdl2009
ViGOR: grouping-based discovery showed more relevancy, faster, and a better user perception over youtube style interface #jcdl2009
is listening to the panel on digital archival at #jcdl2009.
Megan Winget: talking about game performance archives and participation, making me geek out (silently) more than usual. #jcdl2009
Megan Winget: the rank(ed) amateur [ I read this as "rank(led)" user ... hmmm ....] #jcdl2009
Cathy Marshall: "Storage is cheap. Stewardship is not." #jcdl2009
big props to @ccmarshall for making deletion funny. in our house, though, "deletion couture" wins big awards #jcdl2009
Looking at tweets from Joint Conference on Digital LIbraries - #jcdl2009 on twazzup http://bit.ly/38yaM
...design explicitly for preservation; going to have to focus on automated metadata generation. #jcdl2009
Anne Gilliland: primary materials are different than secondary materials (I.e., evidence); types/sources of content multiply;... #jcdl2009
Michael Lesk: we will save everything; won't catalog anything; search will help us find it; attention is the scare resource #jcdl2009
Panel 1: what to preserve in a born-digital world? And we're off .... #jcdl2009
Panel 1: what to preserve in a born-digital world? And we're off .... #jcdl
"Query Parameters for Harvesting Digital Video" -- focus on preserving context (usage, selection) over raw bits #jcdl2009
eating nice brownies at #jcdl2009. watching marchionini talking about video library contexts.
Waiting for the "What should we preserve from a born-digital world?" panel to start, hoping the answer is not "not much". ;) #jcdl2009
Thinking of skipping paper session 5 and going to see the capitol this afternoon... #jcdl
The food is good at #jcdl2009. Lunch was yummy!
@judielaine @band @_cb_ thanks! (and, cool!) good fodder for a little hacking during talks :) #jcdl2009
Attending JCDL conference in Austin TX
@dchud yes, papers on on the USB drive. #jcdl2009
@dchud (re: papers on usb drives) that was the case last year; haven't checked yet #jcdl2009
ah! just realized that the papers must be on the usb drives. is correct? #jcdl2009
wait maybe it was another conference where t-shirts were free... hmm. #jcdl2009
#jcdl2009 t-shirts arent free this year. darn economy.
@pmedward couldnt really see how it was different to having microsoft word open all the time. #jcdl2009
Global Information Gatherer sounds like another general PIM application, Windows-only; why not live in the "cloud" for the campus? #jcdl2009
talk: disambiguating author names w/random forests: http://bit.ly/WREcM
RT @pmedward: Borgman: "Communities are rarely as homogenous as they appear." [amen!] #jcdl2009
is getting the session he is working ready at #jcdl2009.
is getting the session he is working ready at #jcdl2009. I always
@sshreeves I'll be there tomorrow on my way to Austin. Are you going to come over to texas for JCDL?
"interoperability is still a major challenge" - only if you think you're supposed to achieve it yourself, imho #jcdl2009
Borgman: "Communities are rarely as homogenous as they appear." [amen!] #jcdl2009
@vphill Wondering if we need to think of Digital Libraries as activities rather than things? #jcdl2009
Hey @inforama, tonight's (CDT) paper session at JCDL is in Second Life.
So much of the "building capacity in higher ed via technology adoption" argument seems to adopt a "banking model" of education. #jcdl2009
nice to see science commons get a mention #jcdl2009
"global warming" on screen after discussing sensor data, a/c kicks in like a wind tunnel. digital libraries are amazing! #jcdl2009
Borgman: computer science should work as a partner with other disciplines [isn't that what 'informatics' covers? :-/ ] #jcdl2009
@dchud I didn't get the diff between the Huntington Lib collection and the Google search first hits. #jcdl2009
looks to me like the first hit for "shakespeare's first folio" indeed is it... she said it isn't... am i missing something? #jcdl2009
borgman remembers seeing the poster, in the corner of the room, on the google prototype prior to the 1998 brin and page paper #jcdl2009
C Borgman: "digital library" - no longer used much. #jcdl2009
christine borgman steps up for keynote. Ed Fox was very complementary about her, and their time working together in the past. #jcdl2009
been years since i've been to a conf where more wore shirts w/collars &amp; nice pants than wore tshirts &amp; jeans. whatever that means. #jcdl2009
@dchud #jcdl2009 - no wrestling needed.
Wow - two speakers from Asia were not given permission to travel due to the H1N1 virus. #jcdl2009
@gingdottwit easy enough. #jcdl2009 it is!
@dchud #jcdl2009 has many more tweets, despite being the longer tag.
sitting in keynote of JCDL09. well done, says chair, for surviving the 103f heat so far. #jcdl2009
@band so #jcdl2009 or #jcdl09? shall we arm-wrestle over it at the break?
Waiting for Christine Borgman's keynote to start. #jcdl2009
Off to join the #jcdl2009 crowd.
On my way to JCDL
Made it to Austin, TX for JCDL; verdict is: it's hot.
In between flights at DFW on my way to #jcdl2009
@jschneider agreed, PLS RT and help more Iranians to see what's going on at Twitter: http://tr.im/oBom #iranelection #jcdl2009
Updating secondlife in prep for tomorrow's #jcdl2009 poster session. Everything feels trivial, thinking of Iran.
Arrived in Austin for #jcdl2009
anxiety is ramping up again for jcdl. whose bright idea was it for me to give a workshop talk?
JCDL doctoral consortium presentation went pretty well. One more to go!
Really impressed with the food at JCDL!
Sitting in the Oakland airport on the way to JCDL in Austin.
In screening line at SJC, heading to AUS for #JCDL2009 . Do I need my middle name on my ticket?
didn't sleep much last night, and is now paying for in the tutorial he is attending at JCDL.
First day at JCDL, trying to wake up so I can go downstairs and register.
@band looking forward to hearing about JCDL next Sat
[blogged] Poster presentation at JCDL 2009: I'll be leaving for Austin, TX, tomorrow afternoon to atten.. http://snipurl.com/k4lty
JCDL 2009 | Preserving the past, designing the future--today. http://ff.im/-3YMGJ
at the AT&amp;T conf center, U. Austin, Texas, for JCDL 2009
On the way to the airport for Austin, and JCDL
is in the JCDL SV meeting.
Registered for #jcdl2009: http://www.jcdl2009.org/ - looking forward to digital library chit-chat.
On the way to austin for jcdl 2009
slidedeck ready; packing for JCDL in TX. At least the forecast is cooler now(?)... http://bit.ly/TrWRa
is in Austin for JCDL.
@fak3r attending JCDL next week, hanging out with my brother this weekend
back home from ETD 2009 in Pittsburgh, getting ready for JCDL 2009 in Austin
Enjoying that moment of silence between ElPub and JCDL
Ran a 5K at the gym after yesterday's rest day, now (still) finishing JCDL poster, hoping 16pt font is big enough on a 2'x3' poster. (?) :-/
Getting the presentations for #jcdl2009 ready...
plotting out travel and sessions for JCDL in Austin next week. anybody else here going? #jcdl2009
Back in Berkeley doing last minute runs for CLEF and getting ready for JCDL next week. Been playing with Chandler from OSAF - almost useful
Working on slide for 1-min. madness at JCDL; seems to be taking longer than it should, thus no gym today (but will try to run home instead).
Now I am writing tween tweets for #jcdl2009 while at #nmc2009. I don't like it one bit. Would prefer to play.
@jillemery that's this place, right? http://www.meetattexas.com/ same as jcdl? if that's convenient for you, and i can park near, 2pm works!
@jillemery i think monday is totally free - otherwise early evening after jcdl sessions tues or weds works. any of those work?
Joint Conference on Digital Libraries. ACM SIGIR, ACM SIGWEB &amp; IEEE-CS TCDL. Austin, Texas. #librarians. http://www.jcdl2009.org/
