In The New Yorker, Malcom Gladwell argued that online social networks such as Twitter aren’t good for “real” social activism, not least because they support only weak ties. The assumption here is that social activism needs strong ties. In reality, the opposite is true. Mark Granovetter’s classic 1973 paper titled “The Strength of Weak Ties” discussed the relationship between tie strength and social activism. Granovetter considered the redevelopment project of the Italian neighbourhood in Boston in the 60s. The project was widely opposed by the community but went forward. Why? The problem was the absence of weak ties within the Italian neighbourhood. Social life revolved around members and unchanging groups of friends, and the density of strong ties (but relative lack of weak ones) inhibited any political change. Gladwell cited Granovetter’s article but didn’t read it. Gladwell titled his article “Why the revolution will not be tweeted”. Perhaps revolution is not what we need. We might just need people who read what they cite and don’t fall into the trap of “the old dismissing the new” (substitute “telephone” for “twitter”/”facebook” and see how the article reads).#fail
Archive for the ‘social networks’ Category
I’m at the 2010 ECML/PKDD Conference in Barcelona, Spain; I’ll be presenting a paper (slides, paper) at the Workshop on Privacy and Security in Data Mining and Machine Learning (PSDML). Check out the real-time posts on twitter here.
Also, in a similar fashion to what Daniele did for CIKM 2010, here is a
completely arbitrary, title-biased pick of papers from the conference:
- “Demand Drive Tag Recommendation” by G.V. Menezes, J.M. Almeida, F. Belen, M.A. Goncalves, A. Lacerda, E.S. de Moura, G.L. Pappa, A. Veloso, N. Ziviani
- “Fast and Scalable Algorithms for Semi-Supervised Link Prediction on Static and Dynamic Graphs” by R. Raymond and H. Kashima
- “Collecting Traffic Forecasting” by M. Lippi, M. Bertini, P. Frasconi
- “Selecting Information Diffusion Models Over Social Networks for Behavioral Analysis” by K. Saito, M. Kimura, K. Ohara, H. Motoda
- “Finding Critical Nodes for Inhibiting Diffusion of Complex Contagions in Social Networks” by C.J. Kuhlman, V.S.A Kumar, M.V. Marathe, S.S. Ravi, D.J. Rosenkrantz
- “Analysis of Large Multi-Modal Social Networks: Patterns and a Generator” by N. Du, H. Wang, C. Faloutsos
- “Surprising Patterns for the Call Duration Distribution of Mobile Phone Users” by P.O.S. Vez de Melo, L. Akoglu, C. Faloutsos, A.A.F. Loureiro
- “Virus Propagation on Time-Varying Networks: Theory and Immunization Algorithms” by B.A. Prakash, H. Ton, N. Valler, M. Faloutsos, C. Faloutsos
The last four picks were, unsurprisingly, from the social-networks session.
Best Paper Awards:
- “Large Scale Image Annotation: Learning to Rank with Joint Word-Image Embeddings” by J. Weston, S. Bengio, N. Usunier
- “Accelerating Spectral Clustering with Partial Supervision” by D. Mavroeidis
- “A Geometric View of Conjugate Priors” by A. Agarwal and H. Daume III
- “A Game-Theoretic Framework to Identify Overlapping Communities in Social Networks” by W. Chen, Z. Liu, X. Sun, and Y. Wang
Workshops, with some picks:
- Analysis of Complex Networks
- Mining Ubiquitous and Social Environments
- “Towards Adjusting Mobile Devices to User’s Behaviours” F. Jungermann, K. Morik, N. Piatkowski, O. Spinczyk, M. Stolpe, P. Fricke
- “Discovering Trend-Based Clusters in Spatially Distributed Data Streams” by A. Ciampi, A. Appice, D. Malerba
- Supervised and Unsupervised Ensemble Methods and their Applications
- Preference Learning
- “Semantic-Based Destination Suggestion in Intelligent Tourism Information Systems” by M. Ceci, A. Appice, D. Malerba
- Handling Concept Drift in Adaptive Information Systems: Importance, Challenges, Solutions
- Third Generation Data Mining: Towards Service-Oriented Knowledge Discovery
- Privacy and Security Issues in Data Mining and Machine Learning
- “Content Based Filtering in On-line Social Networks” M. Vanetti, E. Binaghi, B. Carminati, M. Carullo, E. Ferrari
- Dynamic Networks and Knowledge Discovery
- “Stream Based Community Discovery via Relational Hypergraph Factorization on Evolving Networks” by C. Bockermann and F. Jungermann
- “Collaboration Based Social Tag Prediction in the Graph of Annotated Web Pages” by H. Rahmani, B. Nobakht, H. Blockeel
- “Automatic Sentiment Monitoring of Specific Topics in the Blogosphere” by F.S. Pimenta, D. Obradovi, R. Schirru, S. Baumann, A. Dengel
- Detection and Identification of Rare Audiovisual Cues
- “Anomaly Detection and Knowledge Transfer in Automatic Sports Video Annotation” by I. Almajai, F. Yan, T. de Campos, A. Khan, W. Christmas, D. Windridge, J. Kittler
Posted in social networks | Comments Closed
“Brave New World of Digital Intimacy” by Clive Thompson on the NYT
“In essence, Facebook users didn’t think they wanted constant, up-to-the-minute updates on what other people are doing. Yet when they experienced this sort of omnipresent knowledge, they found it intriguing and addictive. Why? Social scientists have a name for this sort of incessant online contact. They call it “ambient awareness.” It is, they say, very much like being physically near someone and picking up on his mood through the little things he does — body language, sighs, stray comments — out of the corner of your eye.”
- Researchers say our reaction to social rejection is the same whether it happens online or off
- Elaine Fogel learned this after having a LinkedIn request be turned down
- After experiencing online rejection, Kenneth Loflin altered the way he interacts online
- Some users think you can avoid awkwardness of defriending by sending e-mail to explain
(plus from danah boyd (2006). Friends, Friendsters, and MySpace: ‘defriending’ someone by dropping them from a friend list can result – deliberately or accidentally – in upset feelings )
A recent post here discussed emerging technologies that can be used for advertising on the go- and the threat that they pose to individual privacy. It seems a similar case is now found in online social network sites; places where users volunteer personal information as they interact with their friends. As Daniele mentioned on twitter, a recent TechCrunch article reports on how Facebook now wants to move user information from the private to the public domain (in order to compete with Twitter?)
One of the small steps in doing so involves using your photos to advertise products to your friends:
Facebook occasionally pairs advertisements with relevant social actions from a user’s friends to create Facebook Ads. Facebook Ads make advertisements more interesting and more tailored to you and your friends. These respect all privacy rules. You may opt out of appearing in your friends’ Facebook Ads [..].
Interestingly, until I saw some facebook status updates like this one below, I had no idea:
ATTENTION! FACEBOOK has agreed to let 3rd party advertisers use YOUR posted pictures WITHOUT YOUR PERMISSION! To prevent this: Click on SETTINGS up at the top where you see the Log out link, select PRIVACY,then select NEWS FEEDS & WALL next select the tab that reads FACE BOOK ADS, there is a drop down box, select NO ONE. Then, SAVE your changes. ( RE-POST to let your friends know!)
I’m curious to see what kind of photos will appear, and if facebook measures any change in click-through rates with this feature. However, one of the points this seems to make is that a central aspect of privacy is not only giving users control over the flow of their information, but telling them where it may flow in the first place.
Paper by Johan Bollen, Herbert Van de Sompel, Aric Hagberg, and Ryan Chute
Traditionally, the impact of scientific publications has been expressed in terms of citation counts (e.g., Journal Impact Factor – JIF). Today, new impact measures has been proposed based on social network analysis (e.g., eigenvector centrality) and usage log data (e.g. usage impact factor) to capture scientific impact in the digital era. However, among the plethora of new measures, which is most suitable for measuring scientific impact?
The authors performed a principal component analysis on the rankings produced by 39 different measures of scientific impact. They find that scientific impact is a multi-dimensional construct that cannot be adequately measured by any single indicator, although some are more suitable than others.
From the results, they draw four conclusions that have significant implications on the development of scientific assessment.
- The set of usage measures is more strongly correlated than the set of citation measures, indicating a greater reliability of usage measures calculated from the same usage log data than between citation measures calculated from the same citation data.
- Usage-based measures are stronger indicators of scientific Prestige than many presently available citation measures. Impact factor and journal rank turn out to be strong indicators of scientific Popularity.
- Usage impact measures turn out to be closer to a “consensus ranking” of journals than some common citation measures.
- Contrary to common belief that JIF is the “golden standard”, usage-based measures such as Usage Closeness centrality may be better “consensus” measures than JIF.
The program of SocialCom is out. My picks:
- Deriving Expertise Profiles From Tags (email@example.com)
- Ranking Comments on the Social Web (firstname.lastname@example.org)
- Structure of Heterogeneous Networks (email@example.com)
- Online User Activities Discovery based on Time Dependent Data (firstname.lastname@example.org)
- Evaluating the Impact of Attacks In Collaborative Tagging Environments (email@example.com)
- Community Computing: Comparisons between Rural and Urban Societies using Mobile Phone Data (firstname.lastname@example.org)
(new scientists) “Menezes says he expected communication networks to change during moments of crisis. Yet the researchers found that [in the Enron dataset] the biggest changes actually happened around a month before. For example, the number of active email cliques, defined as groups in which every member has had direct email contact with every other member, jumped from 100 to almost 800 around a month before the December 2001 collapse. ” The title of the paper is “Identification of Organizational Tension Using Complex Networks”. Workshop program
Why I blog about this: Some of us (e.g., Neal) are studying the evolution of social/recommender networks over time. This paper is a nice example of social net evolution analysis put to good use
Posted in social networks | Comments Closed
Peter Bentley of CS UCL is hosting the Royal Institution’s cafe scientifique at 7pm at the RI cafe. Today he’ll discuss: “Is internet-based social networking antisocial?” with Meg Pickard, the Head of Social Media Development, Guardian News & Media, responsible for developing and managing existing and new social web strategy and interactive experiences.
New article by Ross Anderson’s group. It’s beautiful in its simplicity. “Eight Friends are Enough: Social Graph Approximation via Public Listings shows how easy it is for an outsider to work out the structure of friendships on Facebook. (For more, see our blog on Facebook’s technical privacy and its democracy theatre.) ”
In short: Having
- G: undirected graph (e.g., Facebook social net)
- Gk: publicly available portion of G (one in which k outgoing friendship edges have been randomly chosen from G),
they show that the results of applying a certain function f (e.g., centrality, shortest paths, community structure) on Gk are simlar to those of applying f on the entire G! That is, by using the public view (Gk), one is able to infer node centralities, shortest paths, and community structures of the whole G! Scary result for privacy-conscius people! But good news for researchers who need to handle big networks On the scary side, from a partial (public) view of a social network, one is able to guess
- which nodes are central – e.g., 1) marketing companies are able to identify influential individuals and virally spread products through them; or 2) during protests that are self-orginized via text messages, repressive governments are able to identify influential individuals and intercept their text traffic.
- communities – the authors “were ableto divide the [partial] graph into communities nearly as well as using complete graph knowledge.” (Sect 3.5)
Today we had a talk by Cyrus Hall of the University of Lugano. He proposed a p2p sampling algorithm to allow p2p admins to monitor and maintain their nets. He evaluated his algorithm against synthetic p2p topologies, which included kleinberg’s social net model and barabasi’s. For years now, it has been assumed that those traditional models reflect real social nets. However, after the talk, Damon pointed at the existence of exponential random graph models. Traditional models do not assume the knowledge of any global property (e.g., clustering coefficient) of a social network – they do the modeling without any input. By contrast, random graph models are able to model any type of social network on input of the network’s global properties. here is a paper titled “Recent developments in exponential random graph (p*) models for social networks” (pdf). Enjoy!
As part of my recent work on collaborative filtering (CF), I’ve been examining the role that time plays in recommender systems. To date, the most notable use of temporal information (if you’re familiar with the Netflix prize) is that researchers are using time(stamps) to inch their way closer to the million dollar reward. The idea is to use how user-ratings vary according to, for example, the day of the week they were input in order to better predict the probe (and more importantly, the qualifying) datasets. I suppose my only criticism here is that once the million dollars has been won, nobody is going to implement and deploy this aspect of the algorithm (unless you are prepared to update your recommendations every day?) – since, in practice, we do not know when users are going to rate items.
From today’s Business Week:
Why weak ties aren’t always strong. “Researchers at IBM and MIT have found that certain e-mail connections and patterns at work correlate with higher revenue production … they used mathematical formulas to analyze the e-mail traffic, address books, and buddy lists of 2,600 IBM consultants over the course of a year. … They compared the communication patterns with performance, as measured by billable hours. They found that consultants with weak ties to a number of managers produced $98 per month less than average. Why? Those employees may move more slowly as they process “conflicting demands from different managers,” the study’s authors write. They suffer from “too many cooks in the kitchen.”
How to introduce people (matchmaking). They also analyzed methods to introduce employees to colleagues they haven’t yet met (to incent people to participate). … “Geyer and his team are digging for signs of shared interests and behaviors among their colleagues. …In their matchmaking efforts, the IBM team tried a variety of approaches. One used a tool favored by Facebook, recommending friends of common friends. Others analyzed the subjects and themes of employees’ postings on Beehive, words they use, and patents they’ve filed. As expected, some of the systems lined up workers with colleagues they already knew. Others were better at unearthing unknowns. But fewer of them turned out to be good matches. To the frustration of the researchers, some of the workers noted that recommendations looked good, yet they didn’t bother contacting the people. “They put them aside for future reference,” Geyer says. “
There seem to be many reasons why people connect online. For example, on Twitter, I have connected to friends, colleagues, family, people I have met at conferences (or simply know from some of the work), and a couple celebrities (like Tom Waits). These few reasons encompass a largely incomplete list of why two people may connect on a social network; of course, understanding why people connect to each other would give insight into suggesting new connections for people to make… (more…)
Since monetizing from ads wouldn’t work for Twitter (“click through rates on social networks are low – people are there to communicate with each other, not to search for information”), Jeremiah Owyang suggested that Twitter should tap into the lucrative CRM space by offering its own CRM system (or its own analytics system to brands).
That’s not easy, not least because there are unsolved problems that revolve around building a brand management system out of Twitter. The goal of such system would be to make it possible for companies to “monitor, alert, track, prioritize, triage, assign, followup, and report on the interactions with their brands”. So here is a list of cool student projects:
- Build tools for mapping real IDs and pseudonyms (mapping Twitter ID into customer ID –many don’t use their real names)
- Build tools for identifying those people on twitter who influence buying behavior
- Build product recommendation tools that are able to sense and react to users who ask their peers for product recommendations at the point of sale (right in the store).
Useful read: The Facebook Era