Temporal graph patterns by timed automata

Abstract Temporal graphs represent graph evolution over time, and have been receiving considerable research attention. Work on expressing temporal patterns or discovering motifs typically assumes relatively simple constraints, such as journeys or, more generally, existential possibly with finite delays. In this paper we propose to use timed automata express leading a general powerful notion of basic pattern (BGP). The new difficulty is the evaluation constraint large set matchings. An important benefit that they support an iterative state assignment, which can be useful for early detection matches pruning non-matches. We introduce algorithms retrieve all instances BGP match in graph, present results extensive experimental evaluation, demonstrating interesting performance trade-offs. show on-demand algorithm processes total matchings incrementally time preferable when dealing cyclic sparse graphs. On acyclic dense graphs, connectivity partial guaranteed, best achieved by maintaining allowing automaton fully incremental. code datasets used our analysis are available at http://github.com/amirpouya/TABGP .