Time-aware Concept Drift Detection Using the Earth Mover’s Distance

Brockhoff, Tobias; Uysal, Merih Seran; Aalst, Wil M. P. van der

doi:10.1109/icpm49681.2020.00016

Cited by 21 publications

(9 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method proposed in [12] calculates the Earth Mover's Distance (EMD) based on the trace descriptors obtained for sub-logs using a sliding fixed-size window. The approach plots the EMD value in a heatmap showing the values for different window sizes.…”

Section: Visual Analysismentioning

confidence: 99%

“…Regarding the drift perspective, most approaches deal with drifts in the control-flow perspective. The data perspective is explored in [32,59,68], and the time perspective is explored in [4,12,53,62,63,69]. Despite being the most addressed problem, it is difficult to compare the distinct approaches for drift detection in PM.…”

Section: Achievements and Challengesmentioning

confidence: 99%

“…Another vital view in PM is the time perspective, which provides performance indicators as lead, waiting, service, and synchronization times. Despite this perspective being widely explored for process analysis, we found fewer papers exploring it for concept drifts [4,12,53,62,63,69].…”

Section: Perspectives Of Changementioning

confidence: 99%

“…Events recorded with timestamps make it possible to identify bottlenecks or measure service levels [71]. A change in this perspective deals with significant changes in process performance over time [4,12,53,69]. An example can be an activity that is manually executed, e.g., credit card validation, and after a specific date is changed to be automatically performed by the system.…”

Section: Timementioning

confidence: 99%

See 3 more Smart Citations

A Survey on Concept Drift in Process Mining

et al. 2021

View full text Add to dashboard Cite

Concept drift in process mining (PM) is a challenge as classical methods assume processes are in a steady-state, i.e., events share the same process version. We conducted a systematic literature review on the intersection of these areas, and thus, we review concept drift in PM and bring forward a taxonomy of existing techniques for drift detection and online PM for evolving environments. Existing works depict that (i) PM still primarily focuses on offline analysis, and (ii) the assessment of concept drift techniques in processes is cumbersome due to the lack of common evaluation protocol, datasets, and metrics.

show abstract

Section: Visual Analysismentioning

confidence: 99%

Section: Achievements and Challengesmentioning

confidence: 99%

Section: Perspectives Of Changementioning

confidence: 99%

Section: Timementioning

confidence: 99%

See 2 more Smart Citations

A Survey on Concept Drift in Process Mining

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Stertz et al [17] detects the change of event attribute values in business processes in an online setting. Brockho et al [20] applies the earth mover's distance to detect process drifts from both the control-flow and time perspective.…”

Section: Introductionmentioning

confidence: 99%

Detecting Branching Condition Changes in Process Models

Lu¹,

Chen²,

Poon³

2021

Preprint

View full text Add to dashboard Cite

Business processes are continuously evolving in order to adapt to changes due to various factors. One important process drift perspective yet to be investigated is the detection of branching condition changes in the process model. None of the existing process drift detection methods focus on detecting changes of branching conditions in process models. Existing branching condition detection methods do not take changes within the process into account, hence results are inadequate to represent the changes of decision criteria of the process. In this paper, we present a method which can detect branching condition changes in process models. The method takes both process models and event logs as input, and translates event logs into decision sequences for change points detection. The proposed method is evaluated by simulated event logs.

show abstract

A Robust and Accurate Approach to Detect Process Drifts from Event Streams

Chen

Poon

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Time-aware Concept Drift Detection Using the Earth Mover’s Distance

Cited by 21 publications

References 21 publications

A Survey on Concept Drift in Process Mining

A Survey on Concept Drift in Process Mining

Detecting Branching Condition Changes in Process Models

A Robust and Accurate Approach to Detect Process Drifts from Event Streams

Contact Info

Product

Resources

About