A human population communicates through social networks, and such networks appear in a number of different applications of interest: sexual contact networks, criminal networks, web-networks, and more.


Recently the interest in social networks has increased, mainly for two reasons: data from empirical networks are now becoming more and more frequent, and the interest in the area grew quickly after Watts and Strogatz (1999) popularised the area (“You are only 6 handshakes away from the president”).


The present project aims at studying stochastic models for social networks, both static and time-dynamic, which gives desirable properties of the network. Most empirical networks are only partially observed, for example from questionnaire to a sample in the population. In such cases only certain properties of the network can be obtained, for example:
the number of individuals an individual is connected to (the degree distribution), how frequently two connected individuals of an individual are themselves connected (clustering), and if individuals with many/few connections tend to be connected to individuals that themselves have many/few connections (degree correlation).


The present project will define suitable models for such social networks and describe methods for how the models can be used when analysing empirical networks