Hashtags are widely used for communication in online media. As a condensed version
of information they characterize topics and discussions. For their analysis we apply
methods from network science and propose novel tools for tracing their dynamics in
time-dependent data. The observations are characterized by bursty increases and
decreases of hashtag usage. These features can be reproduced with a mechanistic
model.

We build temporal and weighted co-occurence networks from hashtags. On static
snapshots we infer the community structure using customized methods. On temporal
networks we solve the bipartite matching problem of detected communities at
subsequent timesteps by taking into account higher order memory. This results in a
matching protocol that is robust towards temporal fluctuations and instabilities of the
static community detection. The proposed methodology is broadly applicable and its
outcomes reveal the temporal behavior of online topics.

Capturing the Dynamics of Hashtag-Communities [5]

As an observable, we consider the size of the communities in time. We find that the
distributions of gains and losses are fat-tailed indicating occasional, but large and
sudden changes in the usage of hashtags.

We propose a mechanistic model that incorporates a ranking with respect to a prestigescore, consisting of the size (imitation) and the age (recency) of a community. Themodel reproduces the observations to good agreement and offers an explanation forthe observed ranking dynamics and the resulting bursts.

Opinion dynamics are usually based on building constructive consensus within a confidence radius (so called bounded confidence models). We propose a different mechanism of radicalization, where agents can increase their opinion pairwise. Combined with a second threshold, the agreement radius, this mechanism leads to rich dynamics of either consensus or various version of fragmented opinions.

 In online media, the way of communication changed and based on the observations from a dataset of the German election campaign 2017 on twitter, we assume heterogeneously distributed activity of the agents. This ingredient can turn settings that have produced a stable consensus into an unstable scenario. This points towards the important role that the ways of communication play a crucial role in the formation of public opinion.

Poster [10]