Foundations of Graph Representation Learning in the context of Network and Life Sciences

Graph Neural Networks are specific networks for learning on data with the graph structure. Applications of them are
numerous in fields like biology or social networks. Though extremely useful, they suffer from the assumption that graphs
do not change in time. The new frameworks considering the dynamic nature of graphs are being developed, but they
also do not offer a full picture of graph's dynamics. They often consider a narrow range of possible events in the graph's
evolution, such as edge insertion, but often ignore events like node insertions and deletions or changing edge features.
Extension to all types of events would be my research focus. The complete deep learning model that faithfully represents
the real-world data could be applied to predicting the future states of the graph. It also opens a possibility to improve
solving semi-supervised or unsupervised problems. In this research, we would like to investigate how the knowledge about
the future state of the graph can be leveraged to improve performance in tasks concerning its current state. Moreover, it
is worth noting that the ideas associated with representation learning on dynamic graphs can be applied outside of this
domain. Hence insights from improving the temporal Graph Neural Networks could be potentially applied in other areas,
such as dynamic graph generation.