NetSci 2013

From Brede Wiki
Jump to: navigation, search
Event (help)
NetSci 2013
Location: Copenhagen Denmark
Date & time: 2013-06-03 – 2013-06-07


Databases: Twitter WikiCFP
Search: DuckDuckGo Google Bing Twitter

NetSci 2013 is a international scientific school and conference on network science.


[edit] School

Albert-László Barabási did the school associated with the conference. He made an introduction to network science based on his recent book Network Science. Some of the chapters of the book is available from:

The slides are also available from his website:

Examples: (professional network).

Among the topics covered: sparse networks, Metcalf's law (critique: assumes a fully connected network and that the links are equally important), giant component, isolates, clustering coefficient, degree distribution, path length.

Several network software package was also shown, e.g., Gephi and NetworkX.

[edit] Workshops/Satellites

Workshop/satellite meetings took place at the Technical University of Denmark. List of

[edit] Human Behavior and Network Science

Organizers were Janos Kertesz and Rosario N. Mantegna. Among the speakers were Claudio Castellano, Michele Tumminello, Renaud Lambiotte, Shlomo Havlin and:

Dino Pedreschi 
"Mining social behaviour from big data". Data on car travel in Italy: mobility trajectories.[1] Classifier based on role, e.g., "resident", "commuter" or "visitor". Understanding car pooling impact. mobility atlas in Italy: They have access to unique data from OctoTelematics.
Lada Adamic 
The Anatomy of Large Facebook Cascades, about network cascades.[2] Is all cascade small and shallow? Leskovec 2006; Bakshy et al. 2009 (Secondlife); Bakshy et al. 2011 (Twitter). Liben Nowell and Kleinberg 2008 found long cascades in email chain letters. Adamic 2012 et al. 2012 Facebook status update memes. Regarded "exposed", "clicked", "liked", "commented", "reshared". Conclusion: a small fraction of content is highly viral; large distribution can occur organiall or by starting at a very high degree node; ... Audience size is important, rather than demographics for virality.
Jameson Toole 
"The Effect of Job Loss and Individuals' Social Network and Mobility". Call detail record (CDR) from an European country. The researchers looked for large layoff from plant close by reading news. Identied a town with around 15'000 inhabitants with around 1'000 (around 10%) layoff. By analyzing the data they found around 100 people that the researchers were pretty sure were layoff'ed.
Przemyslaw Grabowicz 
"Distinguishing topical and social groups based on common identity and bond theory"
Taha Yasseri 
"Conflicts and Opinion Clashes in Wikipedia". Revert network of Wikipedia; controversiality measure M. [3][4][5]

[edit] Arts, humanities, and complex networks

Denny Vrandečić 
Keynote on Wikidata. Introduced Wikimedia, Wikipedia, Wikimedia Commons. Wikidata example application: Magnus Manske's genology GeneaWiki . Magnus Manske's speech recognition, Wikidata inference, speech generation

Videos from the workshop is available from:

[edit] Conference

[edit] Wednesday

Bernhard Palsson 
a glimpse of genome-scale science: biological networks (-omics network). Topics: stoichiometic matrix (metabolite-by-reaction matrix), M matrix (metabolism), E matrix (expression), O matrix (operons+regulons+stimulons), ME matrix, OME matrix (operons+metabolism+expression). The real cost of sequencing: higher than you think!: The cost of downstream analysis is increasing relative to sequensing.
Kim Sneppen 
Networks with bacteria and other organization, e.g. Lambda phage. A biologic network for counting. A biological network for selecting one out of many (winner-take-all).
Mason A. Porter 
"Cascades and social influence on networks". Given a social contagion how does one distinguish between genuine effect, homophily and environment. Watts Threshold Model. Porter's new model with two types of contagions: normal and "bonus" contagions. Paper behind the talk is "Multi-stage complex contagions" (2013). Topics: peer pressure. "bonus influence". (z1, z2)-regular random graphs, "June Bug"
Shlomo Havlin 
"From single network to network of networks".
Roberta Sinatra 
"Scientific success: the story of your big hit". Citation data from American Physical Society with all papers published in Physical Review. They looked on the paper ("big hit") with the large number of citation (normalized on some way). "Finger law": the timing of the big hit is difficult to predict, productivity affects size of the big hit (more than just by a random effect).
Satyam Mukherjee 
"Novelty, convention and scientific impact" with Brian Uzzi, Michael Stringer, Benjamin Jones. Analyzing of all 17 million papers in the Web of Science database
Raj Kumar Pan 
World citation and collaboration networks: uncovering the role of geography in science
Cesar A. Hidalgo 
"The atlas of cultural development". Data from multiple language versions of Wikipedia on famous persons.
Brian Keegan 
Building the iron cage: the evolution of multidimensional networks in the U.S. Congress and Cartholic Hierarchy. What features predict promotion? U.S. Congress: around 11'000 members. Data from collected by David Cheney.
Pernille Yde 
"Modular networks of word correlations on Twitter". Previous article: "Modular networks of word correlations on Twitter". Twitter-data. Probabilities of word-pairs.
Enys Mones 
"HIV competition dynamics over dynamical sexual networks." Nodes: members of the population; edges: sexual relationships. Two types of infections.
Vittoria Colizza 
"Who's the spreader? Role of children vs. adults in the local transmission and spatial dissemination of epidemics". H1N1 pdm. Regard two population profiles: adults and children and mobility pattern. Metapopulation. Could be tested as Britain had a different school year than the rest of Europe. Could also be tested as there is different ratio of children in, e.g., Europe and Mexico.
Toshihiro Tanizawa 
"A new network visualization method of time series data". Time series modeling with reduces autoregressive model. Use the minimum spanning tree to reduce the network.
Hartmut Lentz 
"unfolding accessibility as a macroscopic approach to temporal networks". Accessibility graph. Data: pig trade, sexual contacts (Rocha 2010), conference contacts Python code on Github.

[edit] Thursday

Noshir Contractor 
"Some assembly required: organizing in the 21st century": Team science. For virtually all scientific fields research is increasingly done in teams, (Wuchyty, Jones, Uzzi), more cited. Data from nanoHUB and MMORPG. Gold farming in MMORPG. nanoHUB is a collaborative online environment in nano science software "Understanding and enabling network dynamics in virtual communities".
Jennifer Neville 
"Supporting statistical hypothesis testing over graphs". An investigation of the distributional characteristics of generative graphy models.... Developed mixed-KPGMs (Kronecker product graph model, Leskovec & Faloutsos, 2007)
Jure Leskovec 
"Deconvolution of networks into communities": community detection. "The strength of weak ties": find weak ties to identify communities". Ground-truth communities: LiveJournal, Friendster, Orkut, YouTube (nodes: 1'138'873; labeled communities: 30'087), DBLP, Amazon. "Communities as 'tiles'" consistent with work of Simmel 1955 on Web of graph affiliations and later Feld 1981. "community-affiliation graph model". A core-periphery structure emerges from the 'tiling' of communities. Code and data available from How is this model different from the model of Mikkel Schmidt?
Aaron Clauset 
"Academic mobility, prestige and hiring networks". Universities are both producers and consumers. He has data on mobility of US academics between universities. minimum rank violation (MRV) ordering and bootstrap testing. Ranks North American universityes with Stanford on the top. He was able to compute the Gini coefficient and show that is a strong "wealth" inequality. "Only 11% faculty swim upstream".
Sue Moon 
"Network topological structures in the Internet and OSNs". She have look on Cyworld, that is a Korean social network service. Cyworld had a very high penetration in South Korea. Analysis on modularity across social network services. Identifying communities in Tuenti community with fake accounts.
Dirk Brockmann 
"Have we been modeling global disease dynamics all wrong?". The articles Strategies for containing an emerging influenze pandemic in Southeast Asia and Strategies for mitigating an influenza pandemic has a high "assume count". Shortest path tree
Lucas Zeub 
Think locally: a local perspective on community structure in networks : conductance.
Richard K. Darst 
"Detectability limit in community detection". Stochastic Block Model. Detectability: Decelle et al.; Modularity Nadakuditi/Newman.
Marta C. Gonzalez 
"A multi-scale multi-cultural study of commuting patterns incorporating digital traces". Commuting and air travel. "Gravity model". 3 numbers: population in the origin, population in the destination, population in all locations closer than the destination. A universal model for mobility and migration patterns. Extended radiation model uses only population and point-of-interest information to predict commuting pattern.

[edit] Friday

Matthew Jackson 
Groups in and between castes in India. MCMC techniques for estimation in ERGM. However with n=30 nodes, there are 2^435 graphs. Related to Bhamidi, Bresler and Sly (2008). Two parameters one on links another on triangle, but "MCMC estimation techniques are inaccurate: How can we compute parameters?" New model: "SERGM". Statistical exponential random graph models <math>P(g) = {\exp[\beta S(g) ]}/{\text{normalization}} </math>. s is a statistics for the graph, e.g., links, cliques. Probability on the statistics rather than the graph: <math>P(s) = {K(s) \exp[\beta s ]}/{\text{normalization}} </math> The model and its estimation can recreate statistics of the graph (e.g., the first and second eigenvalue) from estimation of other statistics. Paper with Arun Chandrasekhar: Tractable and consistent random graph models available on SSRN.
Jordi Bascompte 
"Plant-animal mutualistic networks: the architecture of biodiversity". Biodiversity and nestedness.

[edit] Young investigators

Dashun Wang 
and Chaoming Song and Albert-Laszlo Barabasi: "The science of success". Mathematical model of citation patterns.
Emoke-Agnes Horvat 
and Andreas Spitz: "Blockbusters or art cinema: Which productions write film history? Identifying central films in a multiplex film citation network". Film references network built from IMDb. The type of reference: follows (Godfather, Godfather 2), remake, features (Casablanca featured in another movie), references (Mulholland Drive shows a poster from another film), "edited from" (especially documentary). Finds that "Star Wars" and "King Kong" are the most influential films.
Rene Pfitzner 
"Between preference: quantifying correlations in the topological dynamics of temporal networks". Temporal networks that are "Well-mixed" or "correlated". Length of shortest time-respecting path.
Nicola Perra 
"Random walks and search in time-varying networks".
Lijun Sun 
"Understanding metropolitan collective encounter patterns". "Familiar stranger": a concept of Stanley Milgram presented in his paper The frozen world of familiar strangers. Public smartcard transportation data. Find the other you have encountered more than once. Evidence of paired regularity.
Tiago Peixoto 
"Parsimonious module inference in large networks". Distinguishes between ad-hoc vs. statistical inference for finding modules networks. Stochastic block models. Uses Minimum description length. Claims his intelligent MC scales well and demonstrates an example of a bipartite network of actors and films from IMDb. Has C++ code with the Python graph-tool library.
Enrico Glerean 
Raj Kumar Pan, Jari Saramäki, Mikko Sams: "The brain as a time-varying network: multiplex network related to parallel cognitive processes". 18 million links.
Luis E. C. Rocha 
and Vincent Blondel : "Burst of vertex activation and epidemics in evolving networks". Mathematical model to model temporal burst.

[edit] Ignite talk

Alexander Mantzaris 
Twitter super bow data set. Power cut during the game. Audi vs Oreo. Katz centrality: how far does the tweet travel. Spike in conversation. Sentiment scores and dynamic centrality score combined during a sports event.
Desmond Higham 
How influential is a node in a network? Dynamic Twitter network. continues time ODE solver.
Chiata Poletto 
"How mobility drives pathogen competition in spatially structured populations".
Ingo Scholtes 
"The paths not taken ... Aggregate networks considered harmful". The fallacy of transitivity in collapsed (aggregate) temporal networks. Can we improve graph layout using betweenness preference? "Causality forces". Causality-preserving layout. Data set: MIT RealityMining. Disentangle networks. GitHub IngoScholtes
V. Nicosia 
Phase transition in the economically modelled growth of a cellular nervous system. "What is on a worms mind". Time, space and co-efficiency for modeling C. elegans worm nervous system.
Amy Yu 
"The dynamics of cultural imports". Over 200.000 unique individuuals in least one Wikipedia. This is very related to the material presented by Cesar A. Hidalgo "The atlas of culture development". Women has surpassed men in certain areas such as art and public figures.
Baruch Barzel 
"Universality in network dynamics".
Amitabh Sharma 
"A disease module captures novel candidate genes and pathways fo asthma"
Thilo Gross 
"Engineering localized dynamics in networks". Spectral problems: percolation, ... Eigenvectors are delocalized, but can there not be localized eigenvalues. Yes. when there are symmetric motifs.
M. Garcia-Herranz 
"Using friends as sensors to detect global-scale contagious outbreaks". Uses the Kwak Twitter data set.
Serguei Saavedra 
and Rudolf P. Rohr, Jordi Bascompte: "Structural stability as an integrative approach to study complex dynamical networks"
Michele Starnini 
"Modeling human dynamics of face-to-face interaction networks". SocioPatterns with RFID device. inter-contacts gap times: Burstiness; Duration of conversations: heterogeneity.
Daniel Smilkov 
"Heterogenous susceptibility make networks more vulnerable to epidemic spreading".
Iyad Rahwan 
and Alex Rutherford, Manuel Cerbrian: "You can run, but you can hide: social mobilization in a global manhunt". "12 huors of separation". Limits of social mobilization. Searchability, blendability, findability.

[edit] Participants

Partial list (apart from speakers):

  1. Brian Keegan
  2. Finn Årup Nielsen
  3. Morten Mørup
  4. Sune Lehmann

[edit] References

  1. Understanding the patterns of car travel
  2. How you met me
  3. Value production in a collaborative environment
  4. The most controversial topics in Wikipedia: a multilingual and geographical analysis
  5. Dynamics of conflicts in Wikipedia
Personal tools