Predictive Analytics for Semi-structured Case Oriented Business Processes

Lakshmanan, Geetika T.; Duan, Songyun; Keyser, Paul Τ.; Curbera, Francisco; Khalaf, Rania

doi:10.1007/978-3-642-20511-8_59

Cited by 16 publications

(13 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neither do the authors envision user support through recommendations. Additionally, [11] aims at detecting which document content has an effect on the process outcome, ultimately to provide user recommendations. The process itself however remains rigid and the document structure is pre-defined.…”

Section: Discussion and Related Workmentioning

confidence: 99%

Self-learning Predictor Aggregation for the Evolution of People-Driven Ad-Hoc Processes

Dorn

Marín

Mehandjiev

et al. 2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Contemporary organisational processes evolve with people's skills and changing business environments. For instance, process documents vary with respect to their structure and occurrence in the process. Supporting users in such settings requires sophisticated learning mechanisms using a range of inputs overlooked by current dynamic process systems. We argue that analysing a document's semantics is of uttermost importance to identify the most appropriate activity which should be carried out next. For a system to reliably recommend the next steps suitable for its user, it should consider both the process structure and the involved documents' semantics. Here we propose a self-learning mechanism which dynamically aggregates a process-based document prediction with a semantic analysis of documents. We present a set of experiments testing the prediction accuracy of the approaches individually then compare them with the aggregated mechanism showing a better accuracy.

show abstract

Section: Discussion and Related Workmentioning

confidence: 99%

Self-learning Predictor Aggregation for the Evolution of People-Driven Ad-Hoc Processes

Dorn

Marín

Mehandjiev

et al. 2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…This flexibility makes our work particularly applicable to detecting changes in semi-structured business processes where execution is not necessarily driven by a formal process model, and thus mining a formal process model first in order to compute process changes may not make sense. In particular, our technique could be used to determine when and the degree to which a mined adaptive representation of a semi-structured business process such as the probabilistic graph by Lakshmanan et al [3] should be updated. To the best of our knowledge, this paper represents one of the first attempts at applying concepts from spectral graph analysis to business process change detection.…”

Section: A Similarity Between Mined Process Modelsmentioning

confidence: 99%

“…In this paper we propose a technique to provide this on the basis of computing graph spectra of incoming execution traces. Such information could trigger, for instance an update to an existing mined representation of the business process such as the adaptive probabilistic graph proposed in [3] for capturing semistructured business processes. This graph requires updates each time a new execution trace arrives.…”

Section: Introductionmentioning

confidence: 99%

Detecting changes in a semi-structured business process through spectral graph analysis

Lakshmanan

Keyser

Duan

2011

2011 IEEE 27th International Conference on Data Engineering Workshops

Self Cite

View full text Add to dashboard Cite

Semi-structured business processes are emerging at a rapid pace in industries such as government, insurance, banking and healthcare. The workflows underlying these case-oriented processes are non-deterministic. They are mostly driven by human decision making and content status and they may change frequently depending on factors such as economic conditions, legislative policy changes and technological upgrades. This paper describes a method to detect changes in a running business process by conducting spectral graph analysis of sets of execution traces of the process. This method is beneficial because it does not require mining a process model of the business process, and is consequently independent of any assumptions about the nature of the business process. This makes it particularly applicable to case-oriented semi-structured business processes whose lifecycle is not fully driven by a formal process model. In this paper we present our algorithm for computing graph spectra from business process execution traces, and discuss some initial promising results, as well as exciting ideas generated by this research for future work.

show abstract

“…Subsequently, we determine a dependency between message types and activities dynamically through analysis of execution traces. In a similar effort, Lakshmanan et al [12] describe how an ant colony optimization algorithm learns dependencies of document contents (e.g., the impact of certain values within a message) to predict the flow and outcome of a process. They also apply exponential aging to keep the decision probabilities up to date.…”

Section: Related Workmentioning

confidence: 99%

“…For each tuple, we step through the process model in breadth-first style (lines [4][5][6][7][20][21] and locate the corresponding FlowDirection arc and FlowData annotation (lines [8][9][10][11]19). Once found, we increase the FlowData occurrence value (occ(f d, mt)) using exponentially weighted moving average (EWMA) (lines [12][13][14][15]. At the end, FlowData annotations that are not covered by sequence tuples (tup(msg, act)) receive a lower occurrence value (lines 22-23): Message types that show up for the first time, or messages types that arrive early result in a new FlowData annotation (lines [16][17][18].…”

Section: Self-learning Message Flowsmentioning

confidence: 99%

Supporting Dynamic, People-Driven Processes through Self-learning of Message Flows

Dorn

Dustdar

2011

Notes on Numerical Fluid Mechanics and Multidisciplinary Design

View full text Add to dashboard Cite

Abstract. Flexibility and automatic learning are key aspects to support users in dynamic business environments such as value chains across SMEs or when organizing a large event. Process centric information systems need to adapt to changing environmental constraints as reflected in the user's behavior in order to provide suitable activity recommendations. This paper addresses the problem of automatically detecting and managing message flows in evolving people-driven processes. We introduce a probabilistic process model and message state model to learn message-activity dependencies, predict message occurrence, and keep the process model in line with real world user behavior. Our probabilistic process engine demonstrates rapid learning of message flow evolution while maintaining the quality of activity recommendations.

show abstract

Predictive Analytics for Semi-structured Case Oriented Business Processes

Cited by 16 publications

References 14 publications

Self-learning Predictor Aggregation for the Evolution of People-Driven Ad-Hoc Processes

Self-learning Predictor Aggregation for the Evolution of People-Driven Ad-Hoc Processes

Detecting changes in a semi-structured business process through spectral graph analysis

Supporting Dynamic, People-Driven Processes through Self-learning of Message Flows

Contact Info

Product

Resources

About