Keyword aware influential community search in large attributed graphs

We introduce a novel keyword-aware influential community query KICQ that finds the most communities from an attributed graph, where is defined as closely connected group of vertices having some dominance over other groups with expertise (a set keywords) matching terms (words or phrases). first design facilitates users to issue CS intuitively by using terms, and predicates (AND OR). In this context, we propose word-embedding based similarity model enables semantic search, which substantially alleviates limitations exact keyword search. Next, new influence measure for considers both cohesiveness eliminates need specifying values internal parameters network. Finally, two efficient algorithms searching in large graphs. present detailed experiments case study demonstrate effectiveness efficiency proposed approaches.