Exploiting higher-order dependencies for process analytics
| Autor: | Pavlos Delias, Ioannis Kazanidis |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Nomothetic and idiographic
021110 strategic defence & security studies Sequence Point (typography) Event (computing) Computer science 05 social sciences 0211 other engineering and technologies 02 engineering and technology Data science 0506 political science Theoretical Computer Science Event data Control and Systems Engineering Order (exchange) 050602 political science & public administration Computer Science (miscellaneous) Process analytics Engineering (miscellaneous) Event analysis Social Sciences (miscellaneous) |
| Zdroj: | Kybernetes. 49:1253-1266 |
| ISSN: | 0368-492X |
| DOI: | 10.1108/k-09-2018-0500 |
| Popis: | Purpose The purpose of this paper is to present a methodology to reveal complex events structures in events’ occurrences by analyzing event databases, targeting to systematizing events’ analysis and surpassing the need for idiographic approaches. Design/methodology/approach A process-oriented point of view is enabled by purposeful data transformations, and higher-order dependencies are discovered and exploited to capture the flows among the events. Findings Political events do not follow a linear movement that is implied by a sequence, but they occur in varying patterns that cannot be reflected accurately when assuming only first-order dependencies. Originality/value The methodology suggests a novel way to look and to analyze raw event data and it offers an accessible, practicable and supplementary tool as it does not disturb any of the established relevant research designs, and it does not require any additional data to be applied. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|