See also: IRC log
<Jeanne_> 615.253 is Bobby Martin from State of Tennessee
<Jeanne_> 202.566 is EthanMcmahon
<Jeanne_> +1.509.464 is Rachel Flagg
<Jeanne_> 914.262 Zach Tunis
<Jeanne_> 973.327 is Soon Ae Chun
<scribe> Scribe: DeirdreLee
Jeanne introduces speakers
Carlos Iván Chesñevar, National University of the South, Argentina is first speaker
<Jeanne_> Carlos Ivan Chesnevar--DECIDE 2.0: Intelligent Processing of Citizen Opinions in Social Media (slides can be downloaded at http://cs.uns.edu.ar/~cic/w3c-decide20.zip)
<PhilA> Slides available at http://www.w3.org/egov/wiki/images/9/93/EGOV_Meeting_24_August_2012_DECIDE2.0.pdf
benefits of gov 2.0 tools - new ways of communicating with citizens & new opportunities
<Jeanne_> Slides for all presentations today are at the eGov wiki: http://www.w3.org/egov/wiki/Main_Page
<PhilA> Speaker is Carlos Iv�n Ches�evar (firstname.lastname@example.org) http://cs.uns.edu.ar/~cic
<PhilA> CONICET / Laboratory of Research and Development in Artificial Intelligence
<PhilA> Department of Computer Science and Engineering Universidad Nacional del Sur Bah�a Blanca, Argentina
carlos: describes challenges of
... introduces Decide 2.0 project
<PhilA> For the minutes - Carlos' slides contain the bulk of the detail being presented in readily accessible form
<Jeanne_> Challenge is that citizens post comments and suggestions, but it is hard for government officials to be aware of what these opinions and counter-opinions are
<Jeanne_> The framework is on slide 8 and shows how the information is gathered from Facebook, Twitter, and Windowslive to create a citizens' opinion database that creates a contextually understandable argument of pros and cons for government officials to consider
decide 2.0 Framework utilises tweet sentiment tools
<PhilA> Looking for idea of how accurate the pos/neg/neutral classification is claimed to be, esp. for Twitter
All Tweets that have a common context and sentiment are aggregated to form an argument
<PhilA> Presence of given hashtag seems to be requirement
<PhilA> I don't think my tweet from 27 July would be picked up https://twitter.com/philarcher1/status/228963650344460288
<olyerickson> Current slide: 15
<PhilA> Now on slide 17
PhilA: Marketing companies claim about 70% accuracy with sentiment analysis, Phil can claim up to 96%, not necessarily needing specific hashtags
<Jeanne_> PhilA: Looking at accuracy of gathering tweets that refer to Greece without using the #Greece tag. Do decision makers care how accurate the content is or are they looking for a general view of pros and cons?
main goal was to define opinion trees and counter-arguments
<Jeanne_> All--if you have a question, just enter +q to be recognized.
carlos: not necessarily sentiment analysis
<PhilA> My claim (on behalf of i-sieve Techniologies) of high accuracy in classification of pos/neg and neutral doesn't apply to Twitter where context is absent, only things like blog posts and comments on them
Jeanne_: where is the challenges in relation to close the loop with decision-makers
<PhilA> Paraphrasing Jeanne_ "Do government officials actually take any notice?"
Jeanne_: once the user comments are organised, is there still challenges in getting decision-makers to respond to the feedback
<PhilA> A paper was presented at PMOD about South American use of eGov and transparency, see http://www.w3.org/2012/06/pmod/pmod2012_submission_18.pdf and http://www.w3.org/2012/06/pmod/agenda
carlos: no definite answer to what is the best method in feeding this back to decision-makers, but it can be used to identify critical issues
<BobbiMartin> Great presentation!
<PhilA> Soon Ae Chun, Ph.D
<PhilA> City University of New York
<PhilA> College of Staten Island
<PhilA> Now on slide 3
<PhilA> Now slide 5
<PhilA> Again, slides are informative and capture much of what Soon Ae Chun is saying
SoonAeChun: moving from traditional government to social media-based digital government (gov 2.0)
<PhilA> I think we're on slide 8 now
<PhilA> Slide 9
<PhilA> Slide 10
<PhilA> I heard "standard vocabularies" and my W3C light came on ;-)
<PhilA> Slide 11
<PhilA> Slide 12, related to the work of P. Sobkowicz, M. Kaschesky and G. Bouchard
<olyerickson> Can other people please mute?
<Jeanne_> Phil--you'll have to show me how to do that...
<olyerickson> There is precedence for the type of research discussed in Kavanaugh, but from the financial domain. e.g. correlation btween s.m. activity and market trends
<PhilA> Slide 14 looks at the work of G. Lee and Y. Kwak
<Jeanne_> Interesting olyerickson
<PhilA> This is reminiscent of Tim Davies' 5 stars of open data engagement http://www.timdavies.org.uk/2012/01/21/5-stars-of-open-data-engagement/
<olyerickson> Are these metrics (stars) any different for ANY public engagement?
<olyerickson> Overarching question: for all of these papers, do they focus on their immediate countries/regions, the whole" world or a specific interesting country (e.g. USA, UK, EU countries, etc)
<PhilA> Slide 16, paper by K. Andersen, R. Medaglia and H. Kenriksen
<PhilA> That's the "Doctor, I have diagnosed myself as having disease X, please give me drug Y (I know I'm right, I saw it on Facebook)"
General Services Administration, USA
<PhilA> Key links from Rachel: http://www.howto.gov/web-content/resources/tools/terms-of-service-agreements and http://www.usa.gov/Contact/verify-social-media.shtml
<sandro> "Federal Compatible" Terms of Service --- team of lawyers at GSA which works with vendors
<sandro> eg jurisdiction and indemnification
Rachel_Flagg: terms of service
... helping federal agencies and vendors agree on model agreement on how to use social media tools in an appropriate manner for government
<dickg> Archives included in TOS ?
Rachel_Flagg: allows agencies to
communicate using social media tools in a legally sound
... registery where all agencies register their official federal social media accounts
... citizens can then check if social media accounts are genuine or impersonation
GSA provides the registry and tools for agencies to use
scribe: helps for verification of authentic accounts
<BobbiMartin> The State of TN was one of the first States to create a Social Media policy because our former Governor had a false facebook.
Jeanne_: Has GSA's work in creating a 'safe legal space' changed agencies' adoption of social media?
Rachel_Flagg: Setting up these
centralised programmes validates that these social media tools
are good/useful and can be operated in a safe, legal
... there is now also support offered for specific agencies criteria
... there has also been a culture shift, that social media such as Twitter, has become so prevalant in our lives
<olyerickson> Tsk, tsk...
PhilA: Was there much
... how did GSA go about collecting all the social media channels
<BobbiMartin> I had to an incoming call, that might have caused my issue. Sorry.
Rachel_Flagg: participates in
group that meets up regularly and issues of social media and
verification were discussed regularly
... so there was not much resistance
<Jeanne_> Next W3C eGov Interest Group meetings are 3 September 1600-1730 CST (Hong Kong) and 21 September 0800-0930 PST (Los Angeles)
<olyerickson> +1 to thanks to our speakers, scribe and our leaders!
<PhilA> Next meetings 3 Sept (Eurasian) and 21 Sept (Atlantic)
<Jeanne_> Thanks everyone!