Temporal graphs and time series data have one characteristic in common: they treat "time" as an elementary part of the data model. The ubiquity of the two data structures can be observed in many applications, from financial networks and IoT sensor networks to micro-mobility networks. There, entities and their constantly changing relationships produce huge amounts of time series data, whose connections are described by the graph data. As there is no universally valid data model or analysis options, both data structures are typically analyzed separately or merged with manual effort. %Such use cases would benefit from a hybrid data model and analysis features instead of using isolated silo solutions.
In this PhD work, a hybrid data model and operator concept will be studied that seamlessly combines the expressive power of temporal graphs and time series due to their evolutionary nature. Utilizing a planned operator concept, queries, analyses, and predictions can be performed on such hybrid graphs that are impossible by working on the isolated data structures themselves. Our envisioned flexible concatenation of operators to analytical pipelines enables more complex querying, mining and inference tasks.