May 19th, 2009 by Joshua Weinberger

Jessica Tsai is in San Jose this week, for the 3rd International AAAI Conference on Weblogs and Social Media. Her news story will post shortly over is now posted at destinationCRM.com, but we’ve collected a lightly edited selection of her live-twittering here, which we’ll be updating as the day goes on.

For tweets from other #AAAI / #ICWSM attendees, check the Summize results here.

Mon May 18 16:06    First speaker: Lillian Lee, associate prof @ Cornell U.

Mon May 18 17:36    Next: “Gesundheit! Modeling Contagion thru Facebook News Feed”
Eric Sun, Stanford; Itamar Rosenn, Cameron Marlow & Thomas Lento, Facebook

Mon May 18 18:15    Next: Xiao Wei, “Seeking and Offering Expertise across Categories: A Sustainable Mechanism Works for Baidu Knows.”

Mon May 18 18:51    Mary McGlohon “Community Structure and Information Flow in Usenet”

Mon May 18 20:37    Now: Dominique Cardon,”Does showing off help to make friends? Experimenting a Sociological Game on Self-Exhibition & Social Networks”

Mon May 18 21:11    Next: “What are they blogging about? Personality, Topic and Motivation in Blogs” Alastair J. Gill

Mon May 18 21:39    Next Talk:”A Social Identity Approach to Identify Familiar Strangers in a Social Network” Nitin Agarwal

Mon May 18 22:30 Up Now: “You Are Where You Edit: Locating Wikipedia Contributors Through Edit Histories” Michael Lieberman

Mon May 18 23:09 “CourseRank: A Closed-Community Social System through the Magnifying Glass” Georgia Koutrika

Tue May 19 00:13
“Using Transactional Information to Predict Link Strength in Online Social Networks,” Indika Kahanda

Tue May 19 00:16 Last session of the day: “RevRank: A Fully Unsupervised Algorithm for Selecting the Most Helpful Book Reviews,” Oren Tsur

[The complete tweetstream, after the jump...]

Mon May 18 16:06    First speaker: Lillian Lee, associate prof @ Cornell U.
Mon May 18 16:08    Rainie & Horrigan Pew Survey 2007 — In 2006, 31% of US residents use Internet for gathering or sharing pol’t info (60M+ ppl)
Mon May 18 16:18    Y ppl gather info online? 28% for POV WITHIN comm; 34% to get POV OUTSIDE comm; 28% most sites shared POV; 29% most sites challenged POV
Mon May 18 16:21    “Why can’t we just look at words like “great” and “terrible” to rate a review pos or neg?” Sentiment is an active challenge.
Mon May 18 16:39    “Sentiment analysis. is a v. hard problem.” – Lee; Context matters “Read the book” as book review (pos), movie (neg)
Mon May 18 16:41    Is there really an econ. impact to reputation and reviews? Ex:3.46% of ebay sales attrib to seller’s pos rep. = $1.6 bil in Q4 2000
Mon May 18 16:42    “If you believe opinions are influential, then WHICH opinions are influential? Which stick, propagate, stick in ppl’s minds?”– Lee
Mon May 18 16:49    Social factors boosting helpfulness scores: using “real name”; being from New Jersey (for science books); not being from Guam – Lee
Mon May 18 16:51    Relative to other reviews,helpfulness increase when you conform with avg [Bond & Smith '96] or are “brilliant but cruel”[Amabile '83]
Mon May 18 16:54    To be helpful? “As variance among reviews increases, be slightly above the mean. Except in Japan,..best to be slightly below.” -Lee
Mon May 18 17:00    PDF of Lee’s presentation available online here.
Mon May 18 17:14    More information on Lee’s work, which includes publicly available datasets, here.

Mon May 18 17:36    Next: “Gesundheit! Modeling Contagion thru Facebook News Feed”
Eric Sun, Stanford;
Itamar Rosenn, Cameron Marlow & Thomas Lento, Facebook

Mon May 18 17:38    Eric Sun teaches freshman economics at Stanford, talking today about how ideas diffuse in social network
Mon May 18 17:41    Theory of influentials: to be popular, reach tiny group of influentials; over $1 bil in advertising targets influentials, grow 36% year-over-year
Mon May 18 17:43    Duncan Watts’ theory: “Accidental Influencer” – ideas spread like viruses, either you’re susceptible or not.
Mon May 18 17:48    “Influential nodes are no more likely to trigger cascades than avgerage nodes” – Sun on Duncan Watts’ “accidental influencer” theory

Editors’ Note: Some relevant links by and about Duncan Watts:
Is the Tipping Point Toast?
Marketers spend a billion dollars a year targeting influentials. Duncan Watts says they’re wasting their money. [Clive Thompson, FastCompany, Jan 28, 2008]
Too Big to Fail? How About Too Big to Exist?
[Watts writing in Harvard Business Review]
Duncan Watts on Friendship: I
[BusinessWeek, May 2009]
Duncan Watts on Friendship: II
[BusinessWeek, May 2009]
Six degrees of Kevin Bacon is no urban myth
[Yahoo!'s Yodel Anecdotal blog]
E-mailaholics: How to Tell If You’ve Got The Habit
– Duncan Watts’ research on email attributes [MIT's Technology Review - The Physics arXiv blog]

Mon May 18 17:54    Large clusters not started by “One Guy” – roughly 15% of fans in the biggest cluster of each (Facebook) Page are start points – Sun
Mon May 18 17:59    86.4% of paths of Page diffusion involve at least 4 individuals, compared to 38% in real-life study (Brown and Reingen ’87) – Sun
Mon May 18 18:00    Conclusions: “FB News Feed enables long-lasting chains of diffusion that may reach many more ppl than real-life diffusion chains”-Sun
Mon May 18 18:01    “FB network is v. connected, ideas with good receptiveness will attract wide, long connected clusters” – Sun
Mon May 18 18:03    Long chains are not a function of FB user age, activity, users demo, really just about news feed algorithm(i.e., feed exposure). -Sun
Mon May 18 18:05    Other than early adopter props, no reward yet for users who are chain starters – Sun

Mon May 18 18:15    Next: Xiao Wei, “Seeking and Offering Expertise across Categories: A Sustainable Mechanism Works for Baidu Knows.”
Mon May 18 18:36    Baidu Knows, similar to Yahoo! Answers. From Dec 2007 – May 2008, 9.3 mil Qs asked (w/ some Qs deleted/censored), 5,210,163 resolved
Mon May 18 18:36    Baidu uses points to encourage participation of askers (55.01%) &answerers (22.39%).Those who do both (22.6%) are most active on site

Mon May 18 18:51:37    Mary McGlohon “Community Structure and Information Flow in Usenet”
Mon May 18 20:34:05    “Henri Bendel recently announced they’ll stop selling clothes…merchandise is reportedly heading to TJ Maxx.” http://snurl.com/ia38q

Mon May 18 20:37:42    Now: Dominique Cardon,”Does showing off help to make friends? Experimenting a Sociological Game on Self-Exhibition & Social Networks”
Mon May 18 20:41:16    Play Cardon’s sociological game — Sociogeekhere.
Mon May 18 20:41:58    Warning: To play http://sociogeek.com will require an understanding of French. Apologies for any inconvenience.
Mon May 18 20:49:56    self-exhibition is correlated to age (r=.114); extraversion index (r = .224); & socializing index (r=.14) -Cardon
Mon May 18 20:55:55    no correlation between self-exposition index & SNS use frequency – Cardon
Mon May 18 20:57:56    Self-exposition on Web is a social construction, not an unintended behavior. – Cardon
Mon May 18 21:02:07    Cardon breaks down people into five groups: Modest; Traditional Self-Exhibitionist; Bodily Immodest; Show-off; Provocative
Mon May 18 21:03:49    SNS encourages homophily–prefer ppl w/sim. education & socio-economic status, but allows for more hetero social capital (interests)

Mon May 18 21:11:38    Next: “What are they blogging about? Personality, Topic and Motivation in Blogs” Alastair J. Gill
Mon May 18 21:30:09    General motivations for blogging: exhibitionist, extravert, auto-therapeutic, self-representation, self-promotion, creativity – Gill
Mon May 18 21:34:45    “Blogs unsurprisingly mainly focused on self,” Gill

Mon May 18 21:39:39    Next Talk:”A Social Identity Approach to Identify Familiar Strangers in a Social Network” Nitin Agarwal
Mon May 18 21:42:17    Agarwal talking about “Identifying Familiar Strangers” – People who aren’t in your network, but may share similar interests.
Mon May 18 22:03:04   understanding 1 blogger gives sensible glimpse into others = better customization, personalization and recommendation to niche groups

Mon May 18 22:30:50 Up Now: “You Are Where You Edit: Locating Wikipedia Contributors Through Edit Histories” Michael Lieberman

Mon May 18 23:09:35 “CourseRank: A Closed-Community Social System through the Magnifying Glass” Georgia Koutrika
Mon May 18 23:20:23     CourseRank: educational and social site for Stanford students (Koutrika is postdoc researcher at Stanford http://snurl.com/iansv)
Mon May 18 23:21:09    Smaller communities breed more active users (more intimate?). Philosophy tops list of most active users; Int’t Relations at bottom
Mon May 18 23:26:07   @sechu, [student at] Stanford, [tells CRM] that she relies more on face-to-face advice from upperclassmen rather than CourseRank
Mon May 18 23:27:09    @sechu “I would use it if they had more people update it with the grade they got which is I think the real value add”
Mon May 18 23:29:08   Misperception: Students give unbiased well-informed ratings; high grades = high rating, low grade = low rating.
Mon May 18 23:39:13    80/20 rule applies to SNS. 20% created by intermittent; 80% by active; 90% of ppl on SNS are “lurkers” -Koutrika

Tue May 19 00:13:22
“Using Transactional Information to Predict Link Strength in Online Social Networks,” Indika Kahanda

Tue May 19 00:16:32 Last session of the day: “RevRank: A Fully Unsupervised Algorithm for Selecting the Most Helpful Book Reviews,” Oren Tsur
Tue May 19 00:53:39   Most reviews on Amazon are repetitive, limited contribution, poorly written, and unnoticed – Tsur

Miscellaneous tweets:

Mon May 18 15:51    176 registrations this morning (up slightly/same as 2008), 8 countries, 47% students
Mon May 18 15:56    Nice! Rutgers representation here
Mon May 18 15:58    haha confusion over what hashtag to use. [going with #ICWSM instead of #ICWSM09]
Mon May 18 17:31    There’s *outlets* and free wireless at ICWSM’s AAAI conference. Love conferences in Silicon Valley.
Mon May 18 23:06 Facebook is recruiting at #icwsm

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.



 
RSSFeed

Home | Get CRM Magazine | CRM eWeekly | CRM Topic Centers | CRM Industry Solutions | CRM News | Viewpoints | Web Events | Events Calendar
About destinationCRM | Advertise | Getting Covered | Report Problems | Contact Us
PRIVACY/COOKIES POLICY
Get Adobe Flash player