Abstract

Event processing, analysis, and visualization are increasingly important, and common, problems as telemetry and log recording become ubiquitous. We are still learning about the space of ways to both query and see the results of sequential queries against event-logged data. In this paper, we discuss our Logan event exploration prototype, which is based on regular-expression queries and result histograms; we then discuss the many factors that vary between tools, domains, and datasets — tradeoffs of factors such as session size, cardinality of events, and the presence of event arguments.