A non‐compensatory approach for trace clustering

Delias, Pavlos; Doumpos, Michael; Grigoroudis, Evangelos; Matsatsinis, Nikolaos F.

doi:10.1111/itor.12395

Cited by 16 publications

(17 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…encoding the sequence of events in a case, is then a crucial stage for any PM task [8]. Event logs incorporate multiple information such as activity sequences, time spans, dependency between activities or attribute values, replaceability between activities or resources, concurrent or iterative behavior, and others [3,11,12,19] that can be hardly summarized in a single representation. Encoding transforms this information into a feature space enabling data processing.…”

Section: Introductionmentioning

confidence: 99%

Evaluating Trace Encoding Methods in Process Mining

Barbon

Ceravolo

Damiani

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Encoding methods affect the performance of process mining tasks but little work in the literature focused on quantifying their impact. In this paper, we compare 10 different encoding methods from three different families (trace replay and alignment, graph embeddings, and word embeddings) using measures to evaluate the overlaps in the feature space, the accuracy obtained, and the computational resources (time) consumed with a classification task. Across hundreds of event logs representing four variations of five scenarios and five anomalies, it was possible to identify the edge2vec method as the most accurate and effective in reducing class overlapping in the feature space.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluating Trace Encoding Methods in Process Mining

Barbon

Ceravolo

Damiani

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Quality of the generated clusters should ideally remain consistent for a given technique. We also extended our comparison by measuring the complexity of the identified clusters, through NoHiC, AHC and ActiTraC using various complexity metrics [33], such as: We used a standard configuration of Heuristic Miner [34] for the discovery of the process models from the clusters retrieved by these techniques. Table 10 shows the results of the experiments performed on two different logs.…”

Section: Comparison With Other Trace Clustering Techniquesmentioning

confidence: 99%

Understanding Contrail Business Processes through Hierarchical Clustering: A Multi-Stage Framework

Tariq

Khan²,

Charles

et al. 2020

Algorithms

View full text Add to dashboard Cite

Real-world business processes are dynamic, with event logs that are generally unstructured and contain heterogeneous business classes. Process mining techniques derive useful knowledge from such logs but translating them into simplified and logical segments is crucial. Complexity is increased when dealing with business processes with a large number of events with no outcome labels. Techniques such as trace clustering and event clustering, tend to simplify the complex business logs but the resulting clusters are generally not understandable to the business users as the business aspects of the process are not considered while clustering the process log. In this paper, we provided a multi-stage hierarchical framework for business-logic driven clustering of highly variable process logs with extensively large number of events. Firstly, we introduced a term contrail processes for describing the characteristics of such complex real-world business processes and their logs presenting contrail-like models. Secondly, we proposed an algorithm Novel Hierarchical Clustering (NoHiC) to discover business-logic driven clusters from these contrail processes. For clustering, the raw event log is initially decomposed into high-level business classes, and later feature engineering is performed exclusively based on the business-context features, to support the discovery of meaningful business clusters. We used a hybrid approach which combines rule-based mining technique with a novel form of agglomerative hierarchical clustering for the experiments. A case-study of a CRM process of the UK’s renowned telecommunication firm is presented and the quality of the proposed framework is verified through several measures, such as cluster segregation, classification accuracy, and fitness of the log. We compared NoHiC technique with two trace clustering techniques using two real world process logs. The discovered clusters through NoHiC are found to have improved fitness as compared to the other techniques, and they also hold valuable information about the business context of the process log.

show abstract

“…If a set of numerical representations cannot well represent the running relationship of events in the trace, the similarity between traces in the clustering results may be very low, which ultimately affects the quality of the process model. Researchers explored more methods to improve the result of trace clustering and discover accurate and simple business process models from event logs [9,17,10,30,32].…”

Section: Related Workmentioning

confidence: 99%

A repairing missing activities approach with succession relation for event logs

Liu

Zhang

et al. 2020

Knowl Inf Syst

View full text Add to dashboard Cite

In the field of process mining, it is worth noting that process mining techniques assume that the resulting event logs can not only continuously record the occurrence of events but also contain all event data. However, like in IoT systems, data transmission may fail due to weak signal or resource competition, which causes the company's information system to be unable to keep a complete event log. Based on a incomplete event log, the process model obtained by using existing process mining technologies is deviated from actual business process to a certain degree. In this paper, we propose a method for repairing missing activities based on succession relation of activities from event logs. We use an activity relation matrix to represent the event log and cluster it. The number of traces in the cluster is used as a measure of similarity calculation between incomplete traces and cluster results. Parallel activities in selecting pre-occurrence and post-occurrence activities of missing activities from incomplete traces are considered. Experimental results on real-life event logs show that our approach performs better than previous method in repairing missing activities.

show abstract

A non‐compensatory approach for trace clustering

Cited by 16 publications

References 39 publications

Evaluating Trace Encoding Methods in Process Mining

Evaluating Trace Encoding Methods in Process Mining

Understanding Contrail Business Processes through Hierarchical Clustering: A Multi-Stage Framework

A repairing missing activities approach with succession relation for event logs

Contact Info

Product

Resources

About