Schedule

see paper at Nature Human Behavior 2018

see paper at ICWSM 2020

Spatially embedded social networks have long been at the focus of researchers, but the resolution of most studies was not enough to analyze how connections are formed within large urban areas, and how all this is linked to the socio-economic status of people within the network. First, I am going to show how commuting distance conditions the online social ties of Twitter users in the 50 largest metropolitan areas of the United States. Home and work locations are identified from geolocated tweets that enable us to infer the socio-economic status of individuals. The results suggest that commuting-enabled mixing manifests in reduced levels of income assortativity in online social relationships suggesting a universal role of commuting in integrating disparate social networks in cities. Second, I am going to talk about the difference between the spatial concentration of social ties around people's home and work locations, and how the concentration around the home locations is stronger for people with lower socio-economic status, whereas higher status users have a higher social connection density around their workplaces. These differences between the poor and the rich can also be observed in certain structural metrics of the ego-networks linked to social capital such as the clustering coefficient and the share of supported ties. This suggests that the structure of social networks around homes provide greater social capital for poor users than for rich users.

This talk will cover: (1) The disinformation campaign against Syria's White Helmets. (2) How information seems to be amplified by bots who retweet bots in conversations between China, Pakistan and India, and (3) The messaging about election fraud in the US (work in progress).

The semantic similarity of emotion concepts exhibits both cultural variation and universal structure. In this talk, we demonstrate how colexification networks can be used to operationalize semantic similarity and show how emotion concepts’ semantic similarities differ across language families not only  within emotion concepts, but also in the semantic similarities between cognitive or body concepts and emotion concepts. Finally, we will present a novel network construction method using partial colexifications.

The network theory of mental disorders has gained interest in the recent years. This new framework applies network theory to human behavior and psychiatric disorders by conceiving them as complex systems of mutually influencing components (such as symptoms, of domains of behavior). In this talk, Giovanni Briganti will present an overview of this network theory as well as its statistical counterpart, network analysis, as applied to mental disorders and other psychiatric constructs.

Modern online social networks foster different social bubbles, with some platforms hosting opinions that would not be acceptable on mainstream venues. Gab is one such platform, know as a hub of alt-right politics and users barred from other networks. Gab presents an interesting opportunity for social networks research because near-complete data is available from day one of its creation. In this talk I present our investigation of the evolving user interaction graph, which is viewed both at different times and timescales via a sliding window approach. Varying these timescales, I show that the Gab network is slowly growing on a monthly basis, but smaller windows reveal that this is owing to bursts of arrivals which often follow polarising political events such as suspension of far right users from Twitter or the Charlottesville rally. The network is characterised by interactions between strangers rather than reinforcing links between friends, with around half of the interactions on a daily basis being between users who have never previously interacted. Gab’s connectivity follows the diurnal cycle of the predominantly US and Europe based users, which in the off-peak hours fragments into sub-networks with absolutely no interaction between them. Finally, Gab has a small core of influential users who hold a huge portion of the network’s attention across all timescales, with a much larger pool of users who gain influence only fleetingly.

Social media relies on amplification. It is at the heart of how marginalized communities voice injustices, how elected officials communicate public health guidance, and how misinformation proliferates through vulnerable populations. Each instance of amplification is a networked process emerging from many separate interpersonal interactions around an event, news story, or hashtag. Using the case study of #MeToo, I demonstrate how interactions between those who disclosed early in the hashtag campaign likely reduced the stigma of disclosure, allowing for further amplification of the hashtag. As it continued to be used, these disclosures transcended any single disclosure and coalesced into a larger network, composed of the core participants whose experiences of sexual violence were amplified by a larger periphery of bystanders and other survivors. I argue that this core-periphery structure is a fundamental signature of online amplification, and propose statistical models for how it can be identified empirically in networks. Finally, by applying these models back to the #MeToo case study, I demonstrate their effect on our ability to measure the reach of a hashtag activism event, highlighting the importance of accounting for the core-periphery network structure of amplification.