Enlightening read on @tumblr research by @YahooLabs researchers: arxiv.org/pdf/1403.5206v… [PDF]. Turns out some of my prejudices were wrong.
~RT @YahooLabs: Article from @techreview on new @tumblr research: techre.vu/P8vtyM // My perception of Tumblr: xkcd.com/1025/
“While there’re many #SocialMedia platforms, #BigData research focuses [too much] on #Twitter”-@zeynepparxiv.org/abs/1403.7400T (via@arkaitzz)
@mfouquenet Merci :-)
@moustaki Wow, thanks for the RT. Hadn’t seen this :-)
motherboard Algorithms aren’t only writing stories; they’re breaking them: bit.ly/QnFRDX
@jeffposnick (The YT code’s out-commented.) Share your code? With apiblog.youtube.com/2010/07/new-wa…, how can you get access to the <video> in <iframe>?
Have been playing w/ #WebComponents: hypervideo.herokuapp.com (“View Page Source”, then “Inspect Element”). Great technology!
Some unrelated Facebook photos in recent galleries of @mediagalleries were caused by FB’s non-existent phrase search: support.gnip.com/sources/facebo…
The Evolution of Automated #BreakingNews Stories technologyreview.com/view/525911/th… My work on @mediagalleries in the MIT @techreview!
Twitter Photos are getting more social. We’re unveiling two new features: photo tagging & adding up to 4 photos per Tweet: blog.twitter.com/2014/photos-ju…

mediagalleries #BreakingNews candidate via @WikiLiveMon: nl.wikipedia.org/wiki/FC_Bayern…. Media gallery: pic.twitter.com/mbd2l4YIgM

~RT @mediagalleries: #BreakingNews candidate via @WikiLiveMon: es.wikipedia.org/wiki/Pep_Guard…. Gallery: pic.twitter.com/qJ0VA5FfWM > Congrats, @FCBayern
Knowledge Base Completion via Search-Based Question Answering by West et al. static.googleusercontent.com/media/research… [PDF] #Google #KnowledgeGraph #WWW2014
@mparramon Yeah, this paper and demo triggered the interest of some news media :-)
@LeonDerczynski BLP = en.wikipedia.org/wiki/Wikipedia…. IMHO looks like a smear campaign :-( It’s 2012 “news” after all.
On @Flipboard’s new adaptive magazine layout algorithm: engineering.flipboard.com/2014/03/web-la… (via @paul_irish, CC: @Vjeux @rtroncy @RubenVerborgh)

Shitstorm of @BrendanEich’s #Prop8 donation hits #Wikipedia en.wikipedia.org/wiki/Brendan_E…, pic.twitter.com/cldf30TggU, via @mediagalleries @WikiLiveMon
wsREST The #WSREST2014 workshop program is now online: ws-rest.org/2014/program #WWW2014
The @EFF on @GettyImages’ move allowing to use their images for “free” (free cost, but not free use): eff.org/deeplinks/2014….
sympap #somus2014 program is just out, featuring keynotes by @Benoit_Huet and @fnack and 8 papers. See you at @ICMR2014 ! sites.google.com/site/socialmul…
@prushforth @RubenVerborgh You may follow the thread from here lists.w3.org/Archives/Publi…, this should shine some light on the tweet.
Best(!) explanation of #Hypermedia yet: @RubenVerborgh: [Remove search box on Google] “Everybody knows you just need to type in /?q=query”.
ph_singer Computing Semantic Relatedness from Human Navigational Paths: A Case Study on Wikipedia (IJSWIS Journal) philippsinger.info/ijswis_semanti…
~RT @GoogleCR: [B]ei @GoogleDE & arbeiten daran, besser auf Katastrophen [zu reagieren] > I’m learning to use google.org/crisisresponse… for DE!
@mrowebot Yeah, also loving the raw, honest look :-)
#Turkey, you can still use #Twitter w/ Google’s DNS servers 8.8.8.8, 8.8.4.4. Propagate: pic.twitter.com/8p7x9ZwE2i (via @paulmasonnews @kwelle)

The evolution of #GoogleGlass in only two years: pic.twitter.com/NgCl69fSo7 (via @frischkopp)
mattcutts Ed Snowden spoke (virtually) at #TED. His video: on.ted.com/a0531 and the NSA response: on.ted.com/c04vd
amasad when I said PHP wasn’t so bad at facebook, this is what I meant code.facebook.com/posts/26454483…
Somehow #Chrome v33+ broke the #io13/#io12 slide deck print feature: code.google.com/p/io-2012-slid…. Working in Safari. // CC: @ebidel @Paul_Kinlan
bkeegan “Telling Breaking News Stories from Wikipedia with Social Multimedia.” #arXiv cc @WikiResearch feedly.com/e/NvCFbUlF
My Twitter bots @WikiLiveMon & @mediagalleries are now running on the same Node.js instance→Less traffic for the #Wikipedia API. Winning.
@DocXavi Thanks :-) With Kim (Gabarro) we had briefly considered you for the committee, but too many UPC members already, so never mind.

mediagalleries #BreakingNews candidate via @WikiLiveMon: de.wikipedia.org/wiki/Malaysia-…, pic.twitter.com/lrgoiSDRWQ
Paper on #Facebook’s “human-level” #DeepFace face recognition algorithm by Taigman et al.: facebook.com/download/38828… [PDF]
Telling #BreakingNews Stories from #Wikipedia w/ Social Multimedia: A Case Study of #Sochi2014 on arXiv arxiv.org/abs/1403.4289 @WikiResearch
The Parable of Google Flu: Traps in #BigData Analysis, paper by Lazer et al. on #GoogleFluTrends predictions: uvm.edu/~cdanfort/csc-… [PDF]

jos_tweets Just set this up - nice and fresh! #googletrends #screensaver fplus.me/p/67AV pic.twitter.com/0oyAB4TFHl
MonsieurAZ Registration open to the Web Intelligence Summer School on the Web of Data. ow.ly/upFl0
There’s now a “#RepublicOfCrimea”, unacknowledged by most nations: twitter.com/mediagalleries… (via @WikiLiveMon @mediagalleries) #Crimea
All Apple would need to do is allow people to disable itmss: protocol handling on iOS. @pschiller, please? control.kochava.com/v1/cpi/click?c… (ad!)
Debug #iOS ad crap flow: use (online) cURL (hurl.it), set iOS User-Agent (webapps-online.com/online-tools/u…), GET mob.tictacti.com/dynamic.player…
This #iOS custom URL scheme hijacking crap that auto-opens the App Store has to STOP: mob.tictacti.com/dynamic.player… (real ad(!), opens App Store)
@Olivier_Aubert @_tomchristie Thanks for the pointer to gatling (fefe.de/gatling/)!
@_tomchristie Fair enough ;-) The server was simply the last thing I looked at when debugging… Still banging my head ;-)
#Python’s built-in SimpleHTTPServer (“python -m SimpleHTTPServer”) doesn’t support byte range reqs, so #HTML5 video seeking won’t work. #WTF
#PopcornTime’s gone now, but the code’s still on github.com/popcorn-time/p…. Apparently subapi.com is the central point of failure.
Excellent set of #NodeJS and #npm hints from the good folks over at @heroku: blog.heroku.com/archives/2014/…
Both @WikiLiveMon and @mediagalleries Twitter bots will be down over the weekend due to a server move; wikipedia-irc.herokuapp.com unaffected.
#NodeJS profiling w/ #WebKit #DevTools Agent npmjs.org/package/webkit… and npmjs.org/package/webkit… WebKit DevTools Agent Frontend. Impressed.
The #UNIX “pidof” command for #MacOSX: hints.macworld.com/article.php?st…
timbray Grmf. JSON isn’t actually a subset of JavaScript. timelessrepo.com/json-isnt-a-ja…
mathias Opera 20 disables support for CSP 1.1 paths, to prevent most of @homakov’s cross-domain info leak attacks: blogs.opera.com/security/2014/…
@tcurdt Did you manage to download it?
@MarkusLanthaler Likewise, all the best for your #MarkusPhD defense!
@rtroncy Thanks. However, I guess I won’t make it to Crete.
Finally got a date for the public #PhD defense of #TomsPhD: 2014-05-20 (xkcd.com/1179/ style) in Barcelona UPC Campus, details tbd.
@PFCdgayo #SixWordPeerReview to the rescue ;-) “Described system didn’t change the Web. Reject.” twitter.com/tomayac/status…
#iOS 7.1 supports #AirPlay discovery via #Bluetooth afp548.com/2014/03/10/hid… Au Revoir, #Bonjour nsa.gov/ia/_files/fact…, also lists Bluetooth!
We’re hiring for Sales-Engineer-type #jobs at #Google #Hamburg—or less attractive pitch: be my colleaguegoogle.com/about/careers/…X
@rtroncy Check the last paragraph of chr13.com/2014/03/10/usi…, engineers are already looking at it. Thanks for the pointer, though!
@RubenVerborgh @moustaki Was more aiming at cache expiry… Given today’s mostly static data sources, but then, there’live.dbpedia.orgl0
@RubenVerborgh Great remark, @moustaki! @RubenVerborgh, any insights/recommendations in/for the caching strategy (in-memory & HTTP)?

The current gallery pic.twitter.com/I7C2wozTF1 of @mediagalleries for #MH370 en.wikipedia.org/wiki/Malaysia_… is a quite representative mix of sources.
Finally #Wikidata has a proper reference (and overview) publication: korrekt.org/papers/Wikidat… [PDF] by @vrandezo and @ma_kr.
.@RubenVerborgh: “Towards Web-scale Web querying”: why #SPARQL is n’t the answer & how @LDFragments changes this.
ruben.verborgh.org/blog/2014/03/1…
#Yahoo! rumored to buy @SindiceTech: independent.ie/business/irish…. If true, huge success for the team! (via @ericaxel)
lucas_gonze This post mortem on the Persona project is smart and useful: wiki.mozilla.org/Identity/Perso…
.@DHTMLConf <a href=”#” onclick=”javascript:document.write(‘Looks like #DHTMLconf was all about screens’);”>Click</a> twitpic.com/dxoa6p

philbarker Oh look, YouTube video pages have #schema_org microdata. Have I been missing that for long? google.com/webmasters/too… pic.twitter.com/YUAeunpTx0
@jospang I’m co-running @wsREST and have a paper at #LDOW2014. I might still stop by at SNOW2014… See you in Seoul!
@jospang SNOW2014 looks like one of the more interesting workshops! Reg. travel, I’m flying via Helsinki w/ Finnair. @RevealEU: great blog!
timbray IETF revised the JSON spec. There’s actually some useful stuff in the new RFC7159: tbray.org/ongoing/When/2…
@jospang Reading your Twitter bio (“Media Man”), know my latest project @mediagalleries? Based on breaking news candidates via @WikiLiveMon.
Someone tried to reverse-engineer ranking factors in Google Scholar, calling it Academic #SEO (#ASEO): beel.org/files/papers/2… (via @kwelle)
If I were in charge of @Yahoo!’s PR or if I was @marissamayer, I’d keep the ‘!’, but kill the #yodel. But I’m neither.
#WebDataCommons team have made avail. a #CommonCrawl-based corpus of Web tables of mostly tabular data (not layout) webdatacommons.org/webtables/
@BarnabyWalters Confusing mix of “
\n” in content.html and “\r\n” in content.value. Looks weird, but /me not an XML expert. May be good.
@BarnabyWalters Ah, found it. Parser’s correct, the sample isn’t :-) Search for <article class=”h-entry>, note the missing “. Fixed & works!
Getting started w/ #microformats2 microformats.org/2014/03/05/get…. @BarnabyWalters, is waterpigs.co.uk/php-mf2 100% correct w/ single elem & > 1 type?
The @AtomEditor invite is gone. (Sorry @renepickhardt, it said DM email, repo doesn’t serve. Just request one here atom.io).
I have a spare @AtomEditor invitation. Atom (atom.io) is #GitHub’s new text editor. DM me your email if you want the invite.
jteevan I’ve written a lot about attending conferences with kids. But I can also argue in favor of conferencing alone: slowsearching.blogspot.com/2014/03/attend…
#Yahoo! to stop user access of services w/ #Facebook & #Google IDs: reuters.com/article/2014/0…. @marissamayer’s next big move: #authentication.
googleresearch Call for Participation: Entity Recognition and Disambiguation Challenge! Learn how you can participate at goo.gl/MfEfVi

~RT @mediagalleries: #BreakingNews candidate via @WikiLiveMon en.wikipedia.org/wiki/Carles_Pu…, pic.twitter.com/bLhLEH3qIT #CarlesPuyol to leave #Barça :(
#Wikipedia as a Time Machine—Overview paper on time-aware Wikipedia IR by@stewhirr et al.stewh.com/wp-content/upl…E@WikiResearchh
@dret Restructured? Fsck :( Unsure if helpful, but google.com/about/jobs/sea… (ping me). Return to academia hard? Probably. No @wsREST? Bummer.
@dret Funemployment? No more EMC Corporation? Back to academia? Or just a vacation? Will you attend @wsREST?
@dvanliere Looking at the recent convos on analytics@[…], it looks like the realtime page views API has fallen off the prio list?! Correct?
Large-scale IP geolocation w/ Google BigQuery based on the CC-licensed GeoLite City tables from MaxMind: googlecloudplatform.blogspot.com/2014/03/geoip-…
Hit by “Your system has run out of application memory” issue on #MacOSX #Mavericks: google.com/search?q=%22yo…, discussions.apple.com/thread/5474263… :-(
@informor @MrGreenH OK, still great. Now all I need is a business reason to get me to New York ;-) Thanks for the invitation.
@rtroncy Yay! I was monitoring @MrGreenH’s GitHub repo, so I knew what they were up to ;-) Looking forward to meeting the @seenco folk(s?).
Before #WWW2014, paper at #ICMR2014 workshop #SoMuS2014 Telling #BreakingNews Stories from #Wikipedia w/ Social Multimedia: @mediagalleries.
Awaiting #WWW2014: poster on Bots vs. Wikipedians in #WebSci track, paper on Weaving the Web(VTT) of Data in #LDOW2014 & co-running @wsREST!
RubenVerborgh Awesome! All three #LDOW2014 papers I worked on accepted. Stay tuned for more :-)
thx @tomayac, @natadimou, @Miel_vds, @pietercolpaert!
@abursuc Congratulations! Enjoy the new challenges coming with #AndreisPostdoc!
~RT @bkeegan: A writeup of my analysis and thoughts about whether #Wikipedia can predict #Oscar wins: brianckeegan.com/2014/03/does-w… // Great post!
@rtroncy We are on it. Again, it’s mostly a confidence problem and it’s rather complex for a tweet ;-)
@rtroncy The biggest issue here is confidence. You don’t want to report people dead to $Google_Users_Count until you’re absolutely certain…
#JSON5 is a proposed extension to #JSON that brings #ES5 enhancements to its syntax json5.org (via @dret, @mnot) // Make it so!
@ReaderMeter Looking at what dominates @mediagalleries and @WikiLiveMon, I can only back your observations regarding #Crimea.

#Carnival in #Barcelona making fun of the Spanish crisis—escape of the nation’s brightest brainspic.twitter.com/zKyaxEulfHH :( via@rogerforcadaa
@RubenVerborgh I guess you’re officially Dr. Ruben Verborgh now! Congratulations :-D Now back to business ;-) blog.sgo.to/2014/02/rows.h… by Sam.
ladamic Can cascades be predicted? Upcoming WWW paper with @jcccf, @diojenez, @jure, and Jon Kleinberg: facebook.com/notes/facebook…