Arvutusesteetika ja tööstiili identifitseerimine
Files
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Tänapäeval kasutab meeletu hulk ettevõtteid protsessimudelitel põhinevate äriprotsesside haldamiseks, teostamiseks, monitoorimiseks ja analüüsimiseks protsessiteadlikke infosüsteeme. Lisaks genereerivad need tarkvarasüsteemid monitoorimisetapi osana ka sündmuste logisid, mis kujutavad endast tegelikku faktidest tuletatud (aposteriori) töövoogu ning neid analüüsitakse protsessiandmete hankimise tehnikate abil. Selles töös, osana protsessiandmete hankimisest, tutvustame tööstiili kontseptsiooni töö olemuse kõikehõlmava analüüsi tööriistana. Äriprotsesse ja komponentidevahelist vastastikust sõltuvust saab hinnata tööstiili perspektiivist, mis väljendub meetmetes ja mustrites. Defineerime uuendusliku sündmuste logi esitlemise lähenemise, kus logifaili käsitletakse kujutisena. Lisaks pakume välja meetmete arvutamise ja mustrite identifitseerimise algoritmid, mis põhinevad kujutiste analüüsitehnika ja arvutusesteetika kombinatsioonil. Selle tulemusena on loodud tööstiili hindamise veebipõhise rakenduse prototüüp.
Nowadays, an enormous amount of companies use Process-Aware Information Systems to manage, perform, monitor and analyze business processes based on process models. Moreover, as a part of the monitoring stage, these software systems generate event logs, which represent actual a-posteriori workflow and are analyzed by process mining techniques. In this work, as a part of process mining, we introduce the concept of working style as the tool for comprehensive analysis of the nature of work. Business processes and interdependencies between its constituents can be evaluated from the perspective of working style which is represented by measures and patterns. We define the novel event log representation approach, where the log file is treated as an image. Additionally, we propose measure computation and pattern identification algorithms based on image analysis technique in combination with computational aesthetics. As a result, the web-based prototype application for working style evaluation has been built.
Nowadays, an enormous amount of companies use Process-Aware Information Systems to manage, perform, monitor and analyze business processes based on process models. Moreover, as a part of the monitoring stage, these software systems generate event logs, which represent actual a-posteriori workflow and are analyzed by process mining techniques. In this work, as a part of process mining, we introduce the concept of working style as the tool for comprehensive analysis of the nature of work. Business processes and interdependencies between its constituents can be evaluated from the perspective of working style which is represented by measures and patterns. We define the novel event log representation approach, where the log file is treated as an image. Additionally, we propose measure computation and pattern identification algorithms based on image analysis technique in combination with computational aesthetics. As a result, the web-based prototype application for working style evaluation has been built.