Describing a social network based on a particular type of human social interaction, say, Facebook, is conceptually simple: a set of nodes representing the people involved in such a network, linked by their Facebook connections. But, what kind of network structure would one have if all modes of social interactions between the same people are taken into account and if one mode of interaction can influence another? Here, the notion of a “multiplex” network becomes necessary. Indeed, the scientific interest in multiplex networks has recently seen a surge. However, a fundamental scientific language that can be used consistently and broadly across the many disciplines that are involved in complex systems research was still missing. This absence is a major obstacle to further progress in this topical area of current interest. In this paper, we develop such a language, employing the concept of tensors that is widely used to describe a multitude of degrees of freedom associated with a single entity.
Our tensorial formalism provides a unified framework that makes it possible to describe both traditional “monoplex” (i.e., single-type links) and multiplex networks. Each type of interaction between the nodes is described by a single-layer network. The different modes of interaction are then described by different layers of networks. But, a node from one layer can be linked to another node in any other layer, leading to “cross talks” between the layers. High-dimensional tensors naturally capture such multidimensional patterns of connectivity. Having first developed a rigorous tensorial definition of such multilayer structures, we have also used it to generalize the many important diagnostic concepts previously known only to traditional monoplex networks, including degree centrality, clustering coefficients, and modularity.
We think that the conceptual simplicity and the fundamental rigor of our formalism will power the further development of our understanding of multiplex networks.
This article argues that the embeddedness of interest groups in multiplex networks is an important explanation for variation in interest group influence reputations. Interest groups participate in and learn about the political process through their communication with other groups, collaboration in coalitions, and advocacy in issue areas. As a group engages in communication, collaboration, and issue advocacy, its performance of multiple roles is visible to other interested observers that use this information to make judgments about the group’s contribution (positive or negative) to policy debates. Thus, examining the multiple ways in which interest groups are connected and disconnected helps to account for how their representatives see and think about the community of which they are a part, as well as how they are seen by that community.
Sinan Aral is a leading expert on Social Networks, Social Media and Digital Strategy. He is an Assistant Professor and Microsoft Faculty Fellow at the NYU St...
Among his several lines of research, Dr. Aral has done notable work studying how information diffusion in massive online social networks influences demand patterns, consumer e-commerce behaviors and word of mouth marketing. Sinan is a Phi Beta Kappa graduate of Northwestern University, holds masters degrees from the London School of Economics and Harvard University, and received his PhD from MIT. You can find him on Twitter @sinanaral.
Most models of social contagion take peer exposure to be a corollary of adoption, yet in many settings, the visibility of one's adoption behavior happens through a separate decision process. In online systems, product designers can define how peer exposure mechanisms work: adoption behaviors can be shared in a passive, automatic fashion, or occur through explicit, active sharing. The consequences of these mechanisms are of substantial practical and theoretical interest: passive sharing may increase total peer exposure but active sharing may expose higher quality products to peers who are more likely to adopt.
We examine selection effects in online sharing through a large-scale field experiment on Facebook that randomizes whether or not adopters share Offers (coupons) in a passive manner. We derive and estimate a joint discrete choice model of adopters' sharing decisions and their peers' adoption decisions. Our results show that active sharing enables a selection effect that exposes peers who are more likely to adopt than the population exposed under passive sharing. We decompose the selection effect into two distinct mechanisms: active sharers expose peers to higher quality products, and the peers they share with are more likely to adopt independently of product quality. Simulation results show that the user-level mechanism comprises the bulk of the selection effect. The study's findings are among the first to address downstream peer effects induced by online sharing mechanisms, and can inform design in settings where a surplus of sharing could be viewed as costly.
Ningún funcionario se atreve a decirlo. No hay gobierno municipal, estatal o federal que se decida a llamar a las cosas por su nombre. Pero lo que sucede en Mi
Ningún funcionario se atreve a decirlo. No hay gobierno municipal, estatal o federal que se decida a llamar a las cosas por su nombre. Pero lo que sucede en Michoacán se parece mucho a una guerra civil. Le pueden llamar equis. Pueden pedir suspensión de garantías individuales o disolver los poderes como solicitó el Partido Acción Nacional. Pueden mandar miles de soldados y marinos; de policías federales, estatales y municipales. Pueden militarizar al 100 por ciento el estado, pero lo que sucede en Michoacán se parece a una guerra civil. Y si el gobierno de Enrique Peña Nieto no quiere verlo, si quiere invisibilizarlo, si pretende que miremos a otro lado, comete un grave error. El hecho de ignorar o tapar un problema no lo elimina.
The way information spreads through society has been the focus of intense study in recent years. This work has thrown up…
The way information spreads through society has been the focus of intense study in recent years. This work has thrown up some dramatic results; it explains why some ideas become viral while others do not, why certain individuals are more influential than others and how best to exploit the properties of a network to spread information most effectively.
But today, Chuang Liu at Hangzhou Normal University in China and a few pals have a surprise. They say that when information spreads, there are always blind spots in a network that never receive it. And these unreachable dark corners of the network can be numerous and sizeable.
Until now theorists have predicted that information can always spread until it saturates a network to the point where everybody has received it. These predictions are come from models based on our understanding of diseases and the way they percolate through a population. The basic assumption is that information spreads in the same way.
Node characteristics and behaviors are often correlated with the structure of social networks over time. While evidence of this type of assortative mixing and temporal clustering of behaviors among linked nodes is used to support claims of peer influence and social contagion in networks, homophily may also explain such evidence. Here we develop a dynamic matched sample estimation framework to distinguish influence and homophily effects in dynamic networks, and we apply this framework to a global instant messaging network of 27.4 million users, using data on the day-by-day adoption of a mobile service application and users' longitudinal behavioral, demographic, and geographic data. We find that previous methods overestimate peer influence in product adoption decisions in this network by 300–700%, and that homophily explains >50% of the perceived behavioral contagion. These findings and methods are essential to both our understanding of the mechanisms that drive contagions in networks and our knowledge of how to propagate or combat them in domains as diverse as epidemiology, marketing, development economics, and public health.
Cooperation is essential for successful human societies. Thus, understanding how cooperative and selfish behaviors spread from person to person is a topic of theoretical and practical importance. Previous laboratory experiments provide clear evidence of social contagion in the domain of cooperation, both in fixed networks and in randomly shuffled networks, but leave open the possibility of asymmetries in the spread of cooperative and selfish behaviors. Additionally, many real human interaction structures are dynamic: we often have control over whom we interact with. Dynamic networks may differ importantly in the goals and strategic considerations they promote, and thus the question of how cooperative and selfish behaviors spread in dynamic networks remains open. Here, we address these questions with data from a social dilemma laboratory experiment. We measure the contagion of both cooperative and selfish behavior over time across three different network structures that vary in the extent to which they afford individuals control over their network ties. We find that in relatively fixed networks, both cooperative and selfish behaviors are contagious. In contrast, in more dynamic networks, selfish behavior is contagious, but cooperative behavior is not: subjects are fairly likely to switch to cooperation regardless of the behavior of their neighbors. We hypothesize that this insensitivity to the behavior of neighbors in dynamic networks is the result of subjects’ desire to attract new cooperative partners: even if many of one’s current neighbors are defectors, it may still make sense to switch to cooperation. We further hypothesize that selfishness remains contagious in dynamic networks because of the well-documented willingness of cooperators to retaliate against selfishness, even when doing so is costly. These results shed light on the contagion of cooperative behavior in fixed and fluid networks, and have implications for influence-based interventions aiming at increasing cooperative behavior.
InternetActu.net est un site d'actualité consacré aux enjeux de l'internet, aux usages innovants qu'il permet et aux recherches qui en découlent.
Comment les idées deviennent-elles contagieuses ? La thèse comparant certaines idées à des “virus du cerveau” ne date pas d’hier. Dans son livre Le Gène égoïste, paru en 1976, Richard Dawkins avait créé la notion de mèmes, analogues “mentaux” des gènes, qui étaient capables de s’autorépliquer d’un cerveau à l’autre, et qui, à l’instar des créatures vivantes, cherchaient avant tout à maximiser leur capacité de reproduction. Par la suite, certains avaient essayé de donner corps à une nouvelle science, la mémétique, se basant sur cette notion. L’idée n’a jamais vraiment pris, et peu de chercheurs (à l’exception peut être du philosophe Daniel Dennett et de l’anthropologue des religions Pascal Boyer) ont vraiment continué à travailler sur ces bases. En 2005, le Journal of Memetics fermait définitivement ses portes après huit années d’existence.
En revanche, les mèmes sont devenus un élément constitutif de la pop culture internet. Restait cependant à savoir si cette contagion des idées possède de véritables bases neurales ou si elle n’est rien d’autre… qu’un mème.
Des recherches effectuées à l’UCLA, sous direction du psychologue Matthew Lieberman, donnent aujourd’hui à penser qu’il y aurait une réalité biologique à l’oeuvre dans ce processus de “contamination” intellectuelle. Des chercheurs ont en effet étudié les mécanismes cérébraux impliqués dans le “buzz”.
La publicité en ligne est le moteur économique du web. Pourtant, la faiblesse de l'attention qui lui est accordée conjuguée à la perte de confiance dans les messages promotionnels obligent à repenser la communication en ligne en faveur d'une logique d'influence. Plus de 'earned media' et moins de "paid media". Et pour cause, si beaucoup ne regardent ou ne considèrent pas la publicité, de plus en plus d'internautes s'équipent pour ne pas y être exposés du tout. Selon une étude récente publiée par FairPage, plus de 20 % des internautes utiliseraient AdBlock ou une solution équivalente.
Repenser la place des "top influenceurs"
Cependant, si la publicité n'a cessé d'évoluer, les pratiques en matière de relations influenceurs (ou IRM pour Influencer Relationship Management) sont restées les mêmes depuis plusieurs années. C'est particulièrement vrai pour le ciblage : la nature RP de la discipline encourage à privilégier une sélection des personnes les plus influentes possible selon des critères plus ou moins pertinents. Toucher les leaders d'opinion pour maximiser la résonance en ligne.
La vérité, c'est que cela marche de moins en moins bien. Les stars de la blogosphère sont devenues plus exigeantes que les journalistes des titres les plus lus et prestigieux ; la visibilité attendue reste moindre qu'un passage télé ou un bel article papier et la qualité de la couverture est inégale. Si cette approche conserve une certaine pertinence, elle ne constitue qu'une partie de la réponse.
Mapping healthcare influencers online The network-mapping approach finds its natural home, of course, online, and it is also possible to map the communities that gather around health topics on platforms such as Twitter.
Physician networks tend to reflect real life – publication communities will tend to mirror brick-and-mortar communities such as hospitals – and influencers identified by the literature will probably also hold high-status positions at these institutions. But life online is much more frictionless. In Twitter, we see that communities tend to form around topics and opinions, with factors like geography, offline status and stage presence holding less weight.
For this reason, influence on Twitter is mostly fuelled by the content an individual shares – whether that's defined as the attitudes he or she has, the language he or she uses, or the links he or she shares. Because of this, identifying influencers on Twitter can give us really important insights into what content we should create, and how. These insights ensure our content is likely to be useful to – and shared by – the community we are interested in. They can also, of course, help us understand how to become influential ourselves.
It's easy to see why many of the tools that claim to identify influencers are so well used. They deliver quick, apparently useful, results, and it's only when you dig a little deeper or try to act on the insights that you see how little foundation their recommendations have. But influence is something it's worth taking the time to understand. We believe that robust, quantitative network analytics will revolutionise healthcare marketing in the same way it has revolutionised search engines, social media, and online shopping.
I’ve been doing a lot of thinking about the weight of Klout scores. With the gospel of influencer marketing pollinating the media, it’s hard not to. Like stocks and credit scores, it’s easy to get caught up in the rise and fall of where you stand in relation to others. But is there more to it that makes our Klout monitoring so addicting?
Many a marketing folk will tell you there’s no point in engaging with someone with a Klout score lower than 50. I’m writing today to tell you that’s b-u-l-l. For two reasons:
1) Many a great influencer was born offline.
2) Many a Klout master are just vain social media junkies with nothing of real value to add to the conversation.
In order to identify generic features of online diffusion structure, we study seven diverse examples comprising millions of individual adopters. As opposed to biological contagion, our domain of interest comprises the diffusion of adoptions, where “adop- tion” implies a deliberate action on the part of the adopting individual. In particular, we do not consider mere exposure to an idea or product to constitute adoption. Conta- gious processes such as email viruses, which benefit from accidental or unintentional transmission are therefore excluded from consideration.
Although restricted in this manner, the range of applications that we consider is broad. The seven studies described below draw on different sources of data, were recorded using different technical mechanisms over different timescales, and varied widely in terms of the costliness of an adoption. This variety is important to our con- clusions, as while each individual study no doubt suffers from systematic biases arising from the particular choice of data and methods, collectively they are unlikely to all ex- hibit the same systematic biases. To the extent that we observe consistent patterns across all examples, we expect that our findings should be broadly applicable to other examples of online—and possibly offline—diffusion as well.
The remainder of this paper proceeds as follows. After reviewing the diffusion liter- ature in Section 2, in Section 3 we describe in detail the seven domains we investigate. We present our main results in Section 4, showing that not only are most cascades small and shallow, but also that most adoptions lie in such cascades. In particular, it is rare for adoptions to result from chains of referrals. Finally, in Section 5 we discuss the implications of these results for diffusion models, as well as the apparent discord between our results and the prevalence of popular products, such as Facebook and Gmail, whose success is often attributed to viral propagation.
The latest Onalytica ranking of Big Data influencers on Twitter is now available. The top 200 influencers are listed below.
Methodology: We took the tweets for the last 6 months containing the hashtag #BigData and treated mentions of other tweeters as a link to these. This initial network was made up of 150,486 nodes. After removing smaller, isolated components of the network we calculated mathematical PageRank, which is used to rank the list.
The methodology is different from our previous Big Data ranking which was based on Betweenness Centrality. This methodology weights connectors higher, but we have found that PageRank more effectively highlight the real influencers in the network.
The new entries in the top 200 are more due to people moving from below the 200 water mark to above because they are drawing more attention in the context of Big Data, than due to changes in the methodology.
The influence of influencers is overhyped. We all want to believe that there are these super-hero influencers that can make dramatic changes to organizations, countries, and societies. The idea has...
The idea has been spread in pop-culture in books like Malcolm Gladwell’s the Tipping Point. Recent developments in Network Sciencehave shown that our understanding of influencers – the super-connected individuals in our organizations and society – is more or less wrong.
So what is the truth behind influencers? The science is still figuring it out, but here is what we have learned so far.
It’s all about Micro-Influencers
The super-connected influencer do not exist, instead there are micro-influencers – those that have slightly more influence than the rest of the population influencing those around them to spread their ideas and messages about certain topics. (I would consider my friend Andrew a micro-influencer, he got our whole group of friends drinking high-quality craft beer after he himself jumped into the cult of American craft beer drinking).
We use to think that the human social network was constructed like our airport network (also called scale-free networks), there are hubs in which most traffic can get to most places, thus have huge influence on the flow of information.
The truth is that there are no Chicago O’Hare, or London Heathrow individuals. Why? Because the human network does not work like the airport transportation network. The human capacity to manage relationships is finite. Unlike our major airports, we cannot just construct another terminal in ourselves to deal with more traffic. We have a limited number of relationships we can actively manage and the reach of our direct influence is limited by the relationships we manage.
The average number of friends people have on Facebook is around 200 – but there are some Facebook users who have 2000 friends (the max for an individual account), which is only 1 magnitude greater, not 10 or 20 times greater like we would expect if our human networks were more like airports: like the difference between Colorado Springs Airport traffic and Chicago O’Hare.
Data visualization techniques are enjoying ever greater popularity, notably thank to the recent boom of Big Data and our increased capacity to handle large datasets. Network data visualization tech...
Three tools to visualize personal networks
Data visualization techniques are enjoying ever greater popularity, notably thank to the recent boom of Big Data and our increased capacity to handle large datasets. Network data visualization techniques are no exception. in fact, appealing diagrams of social connections (sociograms) have been at the heart of the field of social network analysis since the 1930s, and have contributed a lot to its success. Today, all this is evolving at unprecedented pace.
In line with these tendencies, the research team of the project ANAMIA (a study of the networks and online sociability of persons with eating disorders, funded by the French ANR) of which I was one of the investigators, have developed new software tools for the visualization of personal network data, with different solutions for the three stages of data collection, analysis, and dissemination of results.
- ANAMIA EGOCENTER is a graphical version of a name generator, to be embedded in a computer-based survey to collect personal network data. It has turned out to be a user-friendly, highly effective interface for interacting and engaging with survey respondents;
Mapping the Spread of #Contagions via Contact Tracing
A contagion passed by human contact, such as SARS or TB, spreads through human networks based on how infectious and susceptible each party is. Multiple contacts with infectious others play a role in the probability of infection. Contagions that flow through human-based networks can be bad(disease, gossip), good(ideas and information) or neutral(money and investments).
Public health officials perform contact tracing to map the spread of the infection and manage its diffusion. The network map above, created at the epidemiology unit of The Centers for Disease Control [CDC], shows the spread of an airborne infectious disease. The map was created using actual contact data from the community in which the outbreak was happening.
Black nodes are persons with clinical disease (and are potentially infectious), pink nodes represent exposed persons with incubating (or dormant) infection and are not infectious, green represent exposed persons with no infection and are notinfectious. The infection status is unknown for the grey nodes.
Unfortunately the 'social butterfly' in this community, the black node in the center of the graph, is also the most infectious -- a super spreader.
Current procedures focus on inoculating the vulnerable -- often the very young and the very old. Network analysis tells us that it may be smarter, and more efficient, to focus on the spreaders -- those with many contacts to many groups.
It is commonly believed that information spreads between individuals like a pathogen, with each exposure by an informed friend potentially resulting in a naive individual becoming infected. However, empirical studies of social media suggest that individual response to repeated exposure to information is significantly more complex than the prediction of the pathogen model. As a proxy for intervention experiments, we compare user responses to multiple exposures on two different social media sites, Twitter and Digg. We show that the position of the exposing messages on the user-interface strongly affects social contagion. Accounting for this visibility significantly simplifies the dynamics of social contagion. The likelihood an individual will spread information increases monotonically with exposure, while explicit feedback about how many friends have previously spread it increases the likelihood of a response. We apply our model to real-time forecasting of user behavior.
The Simple Rules of Social Contagion Nathan O. Hodas, Kristina Lerman
Group emotional contagion, the transfer of moods among people in a group, and its influence on work group dynamics was examined in a laboratory study of managerial decision making using multiple, convergent measures of mood, individual attitudes, behavior, and group-level dynamics. Using a 2 times 2 experimental design, with a trained confederate enacting mood conditions, the predicted effect of emotional contagion was found among group members, using both outside coders' ratings of participants' mood and participants' self-reported mood. No hypothesized differences in contagion effects due to the degree of pleasantness of the mood expressed and the energy level with which it was conveyed were found. There was a significant influence of emotional contagion on individual-level attitudes and group processes. As predicted, the positive emotional contagion group members experienced improved cooperation, decreased conflict, and increased perceived task performance. Theoretical implications and practical ramifications of emotional contagion in groups and organizations are discussed.
All that chatter on social media may be more valuable than we think, say researchers who are mining the postings for clues about how to best control infectious disease.
What people share on social media can sometimes predict the spread of ideas about diseases like the flu, for example, or beliefs about vaccinations. The researchers looked at the social media reactions to issues like childhood immunizations, acceptance of quarantine during the SARS outbreak and public health messages related to infections like influenza.
“If highly connected nodes in the social network (such as celebrities) suggest that the vaccine carries risks, the resulting perception of vaccine risks can propagate quickly through the social network, fueling a vaccine scare and a drop in vaccine coverage,” they write. Cough cough, Jenny McCarthy. But such social connectivity can also help to prevent biological contagion through imitated or culturally promoted behaviors, like covering your mouth when you cough. And control of the SARS virus was largely possible because the general public was accepting of the quarantines.
Social interaction promotes the spread of values, attitudes, and behaviors. Here we report on neural responsesto ideas that are destined to spread. Message communicators were scanned using fMRI during their initial exposure to the to-be-communicated ideas. These message communicators then had the opportunity to spread the messages and their corresponding subjective evaluations to message recipients, outside the scanner. Successful ideas were associated with neural responses in the mentalizing system and the reward system when first heard, prior to spreading them. Similarly, individuals more able to spread their own views to others produced greater mentalizing system activity during initial encoding. Unlike prior social influence studiesthat focus on those being influenced, this investigation focused on the brains of influencers. Successful social influence is reliably associated with an influencer-tobe’sstate of mind when first encoding ideas.
Which thinkers are we guided by? A novel “Thought Leader Map” shows the select group of people with real influence who are setting the trends in the market for ideas. The influencers in philosophy, sociology, economics, and the “hard sciences” have been identified by a Delphi process, asking 50 thought leaders to name their peers. The importance of the influencers is calculated by constructing a cooccurrence network in the Blogosphere. Our main insight is that the era of the great authorities seems to be over. Major thought leaders are rare – the picture is composed of many specialists.
FROM the earliest days of the Internet, robotic programs, or bots, have been trying to pass themselves off as human. Chatbots greet users when they enter an online chat room, for example, or kick them out when they get obnoxious. More insidiously, spambots indiscriminately churn out e-mails advertising miracle stocks and unattended bank accounts in Nigeria. Bimbots deploy photos of gorgeous women to hawk work-from-home job ploys and illegal pharmaceuticals.
Now come socialbots. These automated charlatans are programmed to tweet and retweet. They have quirks, life histories and the gift of gab.
Many of them have built-in databases of current events, so they can piece together phrases that seem relevant to their target audience. They have sleep-wake cycles so their fakery is more convincing, making them less prone to repetitive patterns that flag them as mere programs. Some have even been souped up by so-called persona management software, which makes them seem more real by adding matching Facebook, Reddit or Foursquare accounts, giving them an online footprint over time as they amass friends and like-minded followers.