Conference Paper
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

In service processes, as found in the telecommunications, financial, or healthcare sector, customers compete for the scarce capacity of service providers. For such processes, performance analysis is important and it often targets the time that customers are delayed prior to service. However, this wait time cannot be fully explained by the load imposed on service providers. Indeed, it also depends on resource scheduling protocols, which determine the order of activities that a service provider decides to follow when serving customers. This work focuses on automatically learning resource decisions from events. We hypothesize that queueing information serves as an essential element in mining such protocols and hence, we utilize the queueing perspective of customers in the mining process. We propose two types of mining techniques: advanced classification methods from data mining that include queueing information in their explanatory features and heuristics that originate in queueing theory. Empirical evaluation shows that incorporating the queueing perspective into mining of scheduling protocols improves predictive power.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... If event logs contain resource information, as is the case in the example event log in Table 1, PM can be useful. Some research efforts highlight the activity-resource relationship when mining resource assignment rules (Ly et al. 2006;Liu et al. 2008;Huang et al. 2011;Senderovich et al. 2014a). ...
... Interruptions during a working day can directly be mined from a service log as, e.g., the start of a break is recorded (Senderovich et al. 2014a). However, assuming the presence of a service log limits the applicability of the developed techniques. ...
... 3.2.6, interruptions are logged in service logs (Senderovich et al. 2014a), but assuming its presence limits the applicability of developed techniques. ...
... If event logs contain resource information, as is the case in the example event log in Table 1, PM can be useful. Some research efforts highlight the activity-resource relationship when mining resource assignment rules (Ly et al. 2006;Liu et al. 2008;Huang et al. 2011;Senderovich et al. 2014a). ...
... Interruptions during a working day can directly be mined from a service log as, e.g., the start of a break is recorded (Senderovich et al. 2014a). However, assuming the presence of a service log limits the applicability of the developed techniques. ...
... 3.2.6, interruptions are logged in service logs (Senderovich et al. 2014a), but assuming its presence limits the applicability of developed techniques. ...
Full-text available
Article
The paper focuses on the use of process mining (PM) to support the construction of business process simulation (BPS) models. Given the useful BPS insights that are available in event logs, further research on this topic is required. To provide a solid basis for future work, this paper presents a structured overview of BPS modeling tasks and how PM can support them. As directly related research efforts are scarce, a multitude of research challenges are identified. In an effort to provide suggestions on how these challenges can be tackled, an analysis of PM literature shows that few PM algorithms are directly applicable in a BPS context. Consequently, the results presented in this paper can encourage and guide future research to fundamentally bridge the gap between PM and BPS.
... In such a setting, the policy for handling customers is typically FCFS, within the same class. For derivation of more accurate routing policies, see [10]. These underlying assumptions reflect upon our choices of relevant predictors and parameter estimation techniques throughout the paper. ...
... The experiments for the first call center correspond to the single-class scenario, since we have focused on a single type of customers. For the second call center, three customer types that represent the private sector are considered: VIP, Regular and Low priority (see [10] for further description of the dataset and the priority setting). The synthetic data that we later use for sensitivity analysis comes from a set of simulation runs, based on a multi-class service process. ...
... The first type can be mined via the 'case' perspective in process mining, while the second type can be inferred by applying different queue mining techniques. For example, in [10], resource-scheduling protocols are learned from data and can later be used for delay prediction or simulations of the service process. ...
Article
Information systems have been widely adopted to support service processes in various domains, e.g., in the telecommunication, finance, and health sectors. Information recorded by systems during the operation of these processes provide an angle for operational process analysis, commonly referred to as process mining. In this work, we establish a queueing perspective in process mining to address the online delay prediction problem, which refers to the time that the execution of an activity for a running instance of a service process is delayed due to queueing effects. We present predictors that treat queues as first-class citizens and either enhance existing regression-based techniques for process mining or are directly grounded in queueing theory. In particular, our predictors target multi-class service processes, in which requests are classified by a type that influences their processing. Further, we introduce queue mining techniques that derive the predictors from event logs recorded by an information system during process execution. Our evaluation based on large real-world datasets, from the telecommunications and financial sectors, shows that our techniques yield accurate online predictions of case delay and drastically improve over predictors neglecting the queueing perspective.
... If event logs contain resource information, as is the case in the example event log in Table 1, PM can be useful. Some research efforts highlight the activity-resource relationship when mining resource assignment rules (Ly et al. 2006;Liu et al. 2008;Huang et al. 2011;Senderovich et al. 2014a). ...
... Interruptions during a working day can directly be mined from a service log as, e.g., the start of a break is recorded (Senderovich et al. 2014a). However, assuming the presence of a service log limits the applicability of the developed techniques. ...
... 3.2.6, interruptions are logged in service logs (Senderovich et al. 2014a), but assuming its presence limits the applicability of developed techniques. ...
Conference Paper
This paper focuses on the potential of process mining to support the construction of business process simulation (BPS) models. To date, research efforts are scarce and have a rather conceptual nature. Moreover, publications fail to explicit the complex internal structure of a simulation model. The current paper outlines the general structure of a BPS model. Building on these foundations, modeling tasks for the main components of a BPS model are identified. Moreover, the potential value of process mining and the state of the art in literature are discussed. Consequently, a multitude of promising research challenges are identified. In this sense, the current paper can guide future research on the use of process mining in a BPS context.
... It is clear to see that the LSTMs have drastically suffered through attack A2 for the BPIC2015 logs (AUC drop of more than 20%), while the aggregation encoded models remain relatively stable. There exists literature studying the impact of the resource involved in the execution of a case [20], [22] to the outcome of the case, meaning that permuting this dynamic attribute can have a detrimental effect on the learned behaviour (and performance) of the model. In [22], the authors state that the scheduling of the resource (i.e. which resource is assigned to the case) has an influence on the predictive accuracy of the model. ...
... There exists literature studying the impact of the resource involved in the execution of a case [20], [22] to the outcome of the case, meaning that permuting this dynamic attribute can have a detrimental effect on the learned behaviour (and performance) of the model. In [22], the authors state that the scheduling of the resource (i.e. which resource is assigned to the case) has an influence on the predictive accuracy of the model. However, for the event log BPIC2015, [20] states that the resource involved does not have an impactful influence, as the LTL rule that determines that outcome is rather naive. ...
Full-text available
Conference Paper
As machine and deep learning models are increasingly leveraged in predictive process monitoring, the focus has shifted towards making these models explainable. The successful adoption of a model is dependent on whether decision-makers can trust the predictions and explanations made. However, recent studies have shown that deep learning models are vulnerable to adversarial attacks-small perturbations to the inputs-which trick deep learning algorithms into making incorrect predictions. An additional crucial property is that the explanations are robust against these adversarial attacks when the model decision was not affected. Therefore, this paper introduces a robustness assessment framework by investigating the impact of adversarial attacks on the robustness of predictive accuracy and explanations used in the field of predictive process monitoring. First, adversarial examples of cases in the independent test set are generated to examine the robustness of the predictive model against intentionally manipulated data. Next, the predictive models are compared with similar models trained on data imputed with adversarial attacks. We monitor the impact on predictive performance in terms of AUC at different stages of the case execution. Finally, the robustness of the explanations is calculated as the distance between the original explanations and the explanations extracted from the model trained on attacked data. We test multiple machine and deep learning techniques, namely the transparent logistic regression, random forests with Shapley values, and LSTM neural networks with attention. Results show that especially neural networks suffer from adversarial attacks, and the former two are mostly robust in terms of both predictive accuracy and explanations.
... Nevertheless, in this paper we focus on human resources due to their importance in the execution and management of business processes. (Havur et al., 2015;Senderovich et al., 2014). The literature (see Section 2) outlines distinct techniques and methods that integrate different types of information regarding resource allocation. ...
... Regarding resource allocation, we use an ILP (Schrijver, 1998) approach to solve the problem of identifying the optimal resource designated for allocation from a set of resources that are evaluated according to different metrics. The underlying ILP problem can be posed as follows (de Leoni and van der Aalst, 2013). ...
... -The routing matrix can be inferred by its empirical equivalent, i.e. counts over sums of historical transitions between nodes. -Service policies for routing customers can be discovered using the policy-mining techniques presented in [19]. -The distribution of inter-arrival and service times can be fitted via techniques that were developed and applied in [20,21]. ...
... 8], so that the created models cannot benefit from the analysis techniques developed in Operations Research. In earlier work, therefore, we argued for an explicit representation of the queueing perspective and demonstrated its value for several real-world processes [9,19]. However, the existing techniques all considered the simplistic setting of a single-station system, whereas, this paper addressed the more complex scenario of service processes that are scheduled and have a multi-stage structure that involves resource synchronization. ...
Conference Paper
Service processes, for example in transportation, telecommunications or the health sector, are the backbone of today’s economies. Conceptual models of such service processes enable operational analysis that supports, e.g., resource provisioning or delay prediction. Automatic mining of such operational models becomes feasible in the presence of event-data traces. In this work, we target the mining of models that assume a resource-driven perspective and focus on queueing effects. We propose a solution for the discovery and validation problem of scheduled service processes - processes with a predefined schedule for the execution of activities. Our prime example for such processes are complex outpatient treatments that follow prior appointments. Given a process schedule and data recorded during process execution, we show how to discover Fork/Join networks, a specific class of queueing networks, and how to assess their operational validity. We evaluate our approach with a real-world dataset comprising clinical pathways of outpatient clinics, recorded by a real-time location system (RTLS). We demonstrate the value of the approach by identifying and explaining operational bottlenecks.
... The distribution of interarrival times A t and processing times per class B c can be fitted with the techniques presented in [16,21]. Service policies, P, can be discovered using the policy-mining techniques presented in [22], or assumed to be given, as in the case of discovering a F/J network from a schedule. ...
... In earlier work, therefore, we argued for an explicit representation of the queueing perspective and demonstrated its value for several real-world processes [20,22]. However, the existing techniques all considered the simplistic setting of a singlestation system, whereas, this paper addressed the more complex scenario of service processes that are scheduled and have a multi-stage structure that involves resource synchronization. ...
Article
Service processes, for example in transportation, telecommunications or the health sector, are the backbone of today's economies. Conceptual models of service processes enable operational analysis that supports, e.g., resource provisioning or delay prediction. In the presence of event logs containing recorded traces of process execution, such operational models can be mined automatically.In this work, we target the analysis of resource-driven, scheduled processes based on event logs. We focus on processes for which there exists a pre-defined assignment of activity instances to resources that execute activities. Specifically, we approach the questions of conformance checking (how to assess the conformance of the schedule and the actual process execution) and performance improvement (how to improve the operational process performance). The first question is addressed based on a queueing network for both the schedule and the actual process execution. Based on these models, we detect operational deviations and then apply statistical inference and similarity measures to validate the scheduling assumptions, thereby identifying root-causes for these deviations. These results are the starting point for our technique to improve the operational performance. It suggests adaptations of the scheduling policy of the service process to decrease the tardiness (non-punctuality) and lower the flow time. We demonstrate the value of our approach based on a real-world dataset comprising clinical pathways of an outpatient clinic that have been recorded by a real-time location system (RTLS). Our results indicate that the presented technique enables localization of operational bottlenecks along with their root-causes, while our improvement technique yields a decrease in median tardiness and flow time by more than 20%.
... Erasmus et al. [16] also aim at improving the resource allocation in business processes by considering the resource ability, specified using the Fleishman's taxonomy. As far as performance monitoring is concerned, Senderovich et al. [45] use data mining classification and heuristic methods based on queuing theory to show how the performance of a process can be affected by the scheduling of resources. While focusing on different objectives, these works characterise resources in terms of their expertise or ability, i.e., the set of tasks that they usually perform, workload, i.e., the number of tasks or cases in which they are currently involved in, and the type of outcomes of cases in which they are involved in. ...
Article
Events recorded during the execution of a business process can be used to train models to predict, at run-time, the outcome of each execution of the process (a.k.a. case). In this setting, the outcome of a case may refer to whether a given case led to a customer complaint or not, or to a product return or other claims, or whether a case was completed on time or not. Existing approaches to train such predictive models do not take into account information about the prior experience of the (human) resources assigned to each task in the process. Instead, these approaches simply encode the resource who performs each task as a categorical (possibly one-hot encoded) feature. Yet, the experience of the resources involved in the execution of a case may clearly have an impact on the case outcome. For example, specialized resources or resources who are familiar with a given type of case, are more likely to execute the tasks in a case faster and more effectively, leading to a higher probability of a positive outcome. Motivated by this observation, this article proposes and evaluates a framework to extract features from event logs that capture the experience of the resources involved in a business process. The framework exploits traditional principles from the literature to capture resource experience, such as experiential learning and social ties on the workplace. The proposed framework is evaluated by comparing the performance of state-of-the-art predictive models trained with and without the proposed resource experience features, using publicly available event logs. The results show that the proposed resource experience features may improve the accuracy of predictive models, but that depends on the process execution context, such as the type of process generating an event log or the type of label that is predicted.
... Zhengxing and Huilong presented an approach that mines association rules for resource allocations taking into account the ordered correlations between items in the event log (Zhengxing . The authors of (Senderovich, Weidlich, Gal, & Mandelbaum, 2014a) have presented an approach to mine resource scheduling protocols from event logs. The approach assumes that organisations have a protocol that defines the sequence of activities which are done by a specific set of resources, therefore it is possible to know what activities and what resources will be assigned to those activities if the protocols can be detected. ...
... Process data containing specific information such as interruption events or resource break information (Senderovich et al 2014) can be used directly to inform the aforementioned modeling tasks. However, often such specific information is missing and the interruption time is simply part of the activity duration. ...
... To extend the simulated behaviour, there is a need to increase the capabilities of simulation and discover more complex elements in the simulation model, such as events, decisions, resources and business rules. For such element detection, other approaches, such as BPMN model discovery [8] or resource protocol [38] could be applied, but this needs to be further researched. To clarify the contest of the applicability, there is also a need to investigate data dependency between the data attributes in the event log and the simulation results and how many simulation instances need to be executed to achieve the same expressiveness of the processes as available in the source event logs. ...
Full-text available
Article
There are many approaches on how to analyse business processes, but the simulation is still not widely employed due to high costs associated with simulation model creation. In this paper, an approach on how to automatically generate dynamic business process simulation model is presented. The approach discovers belief network of the process from an event log and uses it to generate a simulation model automatically. Such model then can be further customised to facilitate analysis. For evaluation of the approach, conformance of the simulation results with the source event logs was calculated. The simulation results were event logs that were generated during the simulation of the discovered models. The evaluation showed that the approach could be used for initial simulation model generation.
... Except for (van der Aalst et al. 2003), the papers deal largely with process discovery, and peruse simulation as evidence. Since 2010, the most cited papers all report on formal science or IS engineering, and notably include evaluations Senderovich et al. 2014) or the provision of the research materials and prototypes (Polyvyanyy et al. 2010). ...
Full-text available
Article
The research field of Business Process Management (BPM) has gradually developed as a discipline situated within the computer, management and information systems sciences. Its evolution has been shaped by its own conference series, the BPM conference. Still, as with any other academic discipline, debates accrue and persist, which target the identity as well as the quality and maturity of the BPM field. In this paper, we contribute to the debate on the identity and progress of the BPM conference research community through an analysis of the BPM conference proceedings. We develop an understanding of signs of progress of research presented at this conference, where, how, and why papers in this conference have had an impact, and the most appropriate formats for disseminating influential research in this conference. Based on our findings from this analysis, we provide conclusions about the state of the conference series and develop a set of recommendations to further develop the conference community in terms of research maturity, methodological advance, quality, impact, and progression.
... For example, several approaches have been proposed to predict the remaining processing time of a case depending on characteristics of the partial trace executed [18,22,23]. Other approaches are only targeted to correlating certain predefined characteristics to the process outcome [15,16,24], process service time [25] or the violations of business rules [14]. ...
Full-text available
Article
Process mining can be viewed as the missing link between model-based process analysis and data-oriented analysis techniques. Lion׳s share of process mining research has been focusing on process discovery (creating process models from raw data) and replay techniques to check conformance and analyze bottlenecks. These techniques have helped organizations to address compliance and performance problems. However, for a more refined analysis, it is essential to correlate different process characteristics. For example, do deviations from the normative process cause additional delays and costs? Are rejected cases handled differently in the initial phases of the process? What is the influence of a doctor׳s experience on treatment process? These and other questions may involve process characteristics related to different perspectives (control-flow, data-flow, time, organization, cost, compliance, etc.). Specific questions (e.g., predicting the remaining processing time) have been investigated before, but a generic approach was missing thus far. The proposed framework unifies a number of approaches for correlation analysis proposed in literature, proposing a general solution that can perform those analyses and many more. The approach has been implemented in ProM and combines process and data mining techniques. In this paper, we also demonstrate the applicability using a case study conducted with the UWV (Employee Insurance Agency), one of the largest “administrative factories” in The Netherlands.
Article
Urban mobility impacts urban life to a great extent. To enhance urban mobility, much research was invested in traveling time prediction: given an origin and destination, provide a passenger with an accurate estimation of how long a journey lasts. In this work, we investigate a novel combination of methods from Queueing Theory and Machine Learning in the prediction process. We propose a prediction engine that, given a scheduled bus journey (route) and a 'source/destination' pair, provides an estimate for the traveling time, while considering both historical data and real-time streams of information that are transmitted by buses. We propose a model that uses natural segmentation of the data according to bus stops and a set of predictors, some use learning while others are learning-free, to compute traveling time. Our empirical evaluation, using bus data that comes from the bus network in the city of Dublin, demonstrates that the snapshot principle, taken from Queueing Theory, works well yet suffers from outliers. To overcome the outliers problem, we use Machine Learning techniques as a regulator that assists in identifying outliers and propose prediction based on historical data.
Full-text available
Conference Paper
Process mining is a family of techniques to discover business process models and other knowledge of business processes from event logs. Existing process mining techniques are geared towards discovering models that capture the order of execution of tasks, but not the conditions under which tasks are executed --- also called branching conditions. One existing process mining technique, namely ProM's Decision Miner, applies decision tree learning techniques to discover branching conditions composed of atoms of the form "v op c" where "v" is a variable, "op" is a comparison predicate and "c" is a constant. This paper puts forward a more general technique to discover branching conditions where the atoms are linear equations or inequalities involving multiple variables and arithmetic operators. The proposed technique combine invariant discovery techniques embodied in the Daikon system with decision tree learning techniques.
Full-text available
Article
This paper investigates the possibility of predicting each customer's waiting time in queue before starting service in a multiserver service system with the first-come first-served service discipline, such as a telephone call center. A predicted waiting-time distribution or an appropriate summary statistic such as the mean or the 90th percentile may be communicated to the customer upon arrival and possibly thereafter in order to improve customer satisfaction. The predicted waiting-time distribution may also be used by the service provider to better manage the service system, e.g., to help decide when to add additional service agents. The possibility of making reliable predictions is enhanced by exploiting information about system state, including the number of customers in the system ahead of the current customer. Additional information beyond the number of customers in the system may be obtained by classifying customers and the service agents to which they are assigned. For nonexponential service times, the elapsed service times of customers in service can often be used to advantage to compute conditional-remaining-service-time distributions. Approximations are proposed to convert the distributions of remaining service times into the distribution of the desired customer waiting time. The analysis reveals the advantage from exploiting additional information.
Full-text available
Conference Paper
Companies realize their services by business processes to stay competitive in a dynamic market environment. In particular, they track the current state of the process to detect undesired deviations, to provide customers with predicted remaining durations, and to improve the ability to schedule resources accordingly. In this setting, we propose an approach to predict remaining process execution time, taking into account passed time since the last observed event. While existing approaches update predictions only upon event arrival and subtract elapsed time from the latest predictions, our method also considers expected events that have not yet occurred, resulting in better prediction quality. Moreover, the prediction approach is based on the Petri net formalism and is able to model concurrency appropriately. We present the algorithm and its implementation in ProM and compare its predictive performance to state-of-the-art approaches in simulated experiments and in an industry case study.
Full-text available
Conference Paper
Although simulation is typically considered as relevant and highly applicable, in reality the use of simulation is limited. Many orga- nizations have tried to use simulation to analyze their business processes at some stage. However, few are using simulation in a structured and efiective manner. This may be caused by a lack of training and limita- tions of existing tools, but in this paper we will argue that there are also several additional and more fundamental problems. First of all, the focus is mainly on design while managers would also like to use simu- lation for operational decision making (solving the concrete problem at hand rather than some abstract future problem). Second, there is limited support for using existing artifacts such as historical data and work∞ow schemas. Third, the behavior of resources is modeled in a rather naive manner. This paper focuses on the latter problem. It proposes a new way of characterizing resource availability. The ideas are described and analyzed using CPN Tools. Experiments show that it is indeed possi- ble to capture human behavior in business processes in a much better way. By incorporating better resource characterizations in contemporary tools, business process simulation can flnally deliver on its outstanding promise.
Full-text available
Article
Process mining techniques help organizations discover and analyze business processes based on raw event data. The recently released "Process Mining Manifesto" presents guiding principles and challenges for process mining. Here, the authors summarize the manifesto's main points and argue that analysts should take into account the context in which events occur when analyzing processes.
Full-text available
Article
Telephone call centers are an integral part of many businesses, and their economic role is significant and growing. They are also fascinating sociotechnical systems in which the behavior of customers and employees is closely intertwined with physical performance measures. In these environments traditional operational models are of great value—and at the same time fundamentally limited—in their ability to characterize system performance. We review the state of research on telephone call centers. We begin with a tutorial on how call centers function and proceed to survey academic research devoted to the management of their operations. We then outline important problems that have not been addressed and identify promising directions for future research.
Full-text available
Book
An exploration of concepts and techniques necessary for diagnosing and correcting the problems that create queues.
Full-text available
Conference Paper
It is vital to use accurate models for the analysis, design, and/or control of business processes. Unfortunately, there are often important discrepancies between reality and models. In earlier work, we have shown that simulation models are often based on incorrect assumptions and one example is the speed at which people work. The “Yerkes-Dodson Law of Arousal” suggests that a worker that is under time pressure may become more efficient and thus finish tasks faster. However, if the pressure is too high, then the worker’s performance may degrade. Traditionally, it was difficult to investigate such phenomena and few analysis tools (e.g., simulation packages) support workload-dependent behavior. Fortunately, more and more activities are being recorded and modern process mining techniques provide detailed insights in the way that people really work. This paper uses a new process mining plug-in that has been added to ProM to explore the effect of workload on service times. Based on historic data and by using regression analysis, the relationship between workload and services time is investigated. This information can be used for various types of analysis and decision making, including more realistic forms of simulation.
Full-text available
Conference Paper
Process-aware Information Systems typically log events (e.g., in transaction logs or audit trails) related to the actual business process executions. Proper analysis of these execution logs can yield important knowledge that can help organizations to improve the quality of their services. Starting from a process model, which can be discovered by conventional process mining algorithms, we analyze how data attributes influence the choices made in the process based on past process executions. Decision mining, also referred to as decision point analysis, aims at the detection of data dependencies that affect the routing of a case. In this paper we describe how machine learning techniques can be leveraged for this purpose, and we present a Decision Miner implemented within the ProM framework. KeywordsBusiness Process Intelligence–Process Mining–Petri Nets–Decision Trees
Full-text available
Article
Process mining allows for the automated discovery of process models from event logs. These models provide insights and enable various types of model-based analysis. This paper demonstrates that the discovered process models can be extended with information to predict the completion time of running instances. There are many scenarios where it is useful to have reliable time predictions. For example, when a customer phones her insurance company for information about her insurance claim, she can be given an estimate for the remaining processing time. In order to do this, we provide a configurable approach to construct a process model, augment this model with time information learned from earlier instances, and use this to predict e.g. the completion time. To provide meaningful time predictions we use a configurable set of abstractions that allow for a good balance between "overfitting" and "underfitting". The approach has been implemented in ProM and through several experiments using real-life event logs we demonstrate its applicability.
Full-text available
Article
In a recent paper we introduced the queue-and-idleness-ratio (QIR) family of routing rules for many- server service systems with multiple customer classes and server pools. A newly available server next serves the customer from the head of the queue of the class (from among those he is eligible to serve) whose queue length most exceeds a specified proportion of the total queue length. Under fairly general conditions, QIR produces an important state-space collapse as the total arrival rate and the numbers of servers increase in a coordinated way. That state-space collapse was previously used to delicately balance service levels for the different customer classes. In this sequel, we show that a special version of QIR stochastically minimizes convex holding costs in a finite-horizon setting when the service rates are restricted to be pool-dependent. Under additional regularity conditions, the special version of QIR reduces to a simple policy: Linear costs produce a priority-type rule, in which the least-cost customers are given low priority. Strictly convex costs (plus other regularity conditions) produce a many-server analogue of the generalized-c" (Gc") rule, under which a newly available server selects a customer from the class experiencing the greatest marginal cost at that time.
Full-text available
Article
Motivated by interest in making delay announcements to arriving customers who must wait in call centers and related service systems, we study the performance of alternative real-time delay estimators based on recent customer delay experience. The main estimators considered are: (i) the delay of the last customer to enter service (LES), (ii) the delay experienced so far by the customer at the head of the line (HOL), and (iii) the delay experienced by the customer to have arrived most recently among those who have already completed service (RCS). We compare these delay-history estimators to the standard estimator based on the queue length (QL), commonly used in practice, which requires knowledge of the mean interval between successive service completions in addition to the queue length. We characterize performance by the mean squared error (MSE). We do analysis and conduct simulations for the standard GI=M=s multi- server queueing model, emphasizing the case of large s. We obtain analytical results for the conditional distribution of the delay given the observed HOL delay. An approximation to its mean value serves as a reflned estimator. For all three candidate delay estimators, the MSE relative to the square of the mean is asymptotically negligible in the many-server and classical heavy-tra-c limiting regimes.
Full-text available
Article
Motivated by telephone call centers, we study large-scale service systems with multiple customer classes and multiple agent pools, each with many agents. To minimize sta-ng costs subject to service-level constraints, where we delicately balance the service levels (SLs) of the difierent classes, we propose a family of routing rules called Fixed-Queue-Ratio (FQR) rules. With FQR, a newly available agent next serves the customer from the head of the queue of the class (from among those he is eligible to serve) whose queue length most exceeds a specifled proportion of the total queue length. The proportions can be set to achieve desired SL targets. The FQR rule achieves an important state-space collapse (SSC) as the total arrival rate increases, in which the individual queue lengths evolve as flxed proportions of the total queue length. In the current paper we consider a variety of service-level types and exploit SSC to construct asymptotically optimal solutions for the sta-ng-and-routing problem. The key assumption in the current paper is that the service rates depend only on the agent pool.
Conference Paper
Information systems have been widely adopted to support service processes in various domains, e.g., in the telecommunication, finance, and health sectors. Recently, work on process mining showed how management of these processes, and engineering of supporting systems, can be guided by models extracted from the event logs that are recorded during process operation. In this work, we establish a queueing perspective in operational process mining. We propose to consider queues as first-class citizens and use queueing theory as a basis for queue mining techniques. To demonstrate the value of queue mining, we revisit the specific operational problem of online delay prediction: using event data, we show that queue mining yields accurate online predictions of case delay.
Article
Information systems have been widely adopted to support service processes in various domains, e.g., in the telecommunication, finance, and health sectors. Information recorded by systems during the operation of these processes provide an angle for operational process analysis, commonly referred to as process mining. In this work, we establish a queueing perspective in process mining to address the online delay prediction problem, which refers to the time that the execution of an activity for a running instance of a service process is delayed due to queueing effects. We present predictors that treat queues as first-class citizens and either enhance existing regression-based techniques for process mining or are directly grounded in queueing theory. In particular, our predictors target multi-class service processes, in which requests are classified by a type that influences their processing. Further, we introduce queue mining techniques that derive the predictors from event logs recorded by an information system during process execution. Our evaluation based on large real-world datasets, from the telecommunications and financial sectors, shows that our techniques yield accurate online predictions of case delay and drastically improve over predictors neglecting the queueing perspective.
Article
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, ***, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.
Conference Paper
Discovering predictive models for run-time support is an emerging topic in Process Mining research, which can effectively help optimize business process enactments. However, making accurate estimates is not easy especially when considering fine-grain performance measures (e.g., processing times) on a complex and flexible business process, where performance patterns change over time, depending on both case properties and context factors (e.g., seasonality, workload). We try to face such a situation by using an ad-hoc predictive clustering approach, where different context-related execution scenarios are discovered and modeled accurately via distinct state-aware performance predictors. A readable predictive model is obtained eventually, which can make performance forecasts for any new running process case, by using the predictor of the cluster it is estimated to belong to. The approach was implemented in a system prototype, and validated on a real-life context. Test results confirmed the scalability of the approach, and its efficacy in predicting processing times and associated SLA violations.
Article
With the rise of electronic integration between organizations, the need for a precise specification of interaction behavior increases. Information systems, replacing interaction previously carried out by humans via phone, faxes and emails, require a precise specification for handling all possible situations. Such interaction behavior is described in process choreographies. While many proposals for choreography languages have already been made, most of them fall into the category of interconnection models, where the observable behavior of the different partners is described and then related via message flow. As this article will show, this modeling approach fails to support fundamental design principles of choreographies and typically leads to modeling errors. This motivates an alternative modeling style, namely interaction modeling, for overcoming these limitations. While the main concepts are independent of a concrete modeling language, iBPMN is introduced as novel interaction modeling language. Formal execution semantics are provided and a comprehensive toolset implementing the approach is presented.
Article
This paper introduces a novel methodology for generating scheduling rules using a data-driven approach. We show how to use data mining to discover previously unknown dispatching rules by applying the learning algorithms directly to production data. This approach involves preprocessing of historic scheduling data into an appropriate data file, discovery of key scheduling concepts, and representation of the data mining results in a way that enables its use for job scheduling. We also consider how by using this new approach unexpected knowledge and insights can be obtained, in a manner that would not be possible if an explicit model of the system or the basic scheduling rules had to be obtained beforehand. All of our results are illustrated via numerical examples and experiments on simulated data.
Service Management: Operations, Strategy, Information technology
  • J A Fitzsimmons
  • M J Fitzsimmons
  • J.A. Fitzsimmons
Fitzsimmons, J.A., Fitzsimmons, M.J.: Service Management: Operations, Strategy, Information technology. McGraw-Hill/Irwin Boston (2004)
Multi-Level Workforce Planning in Call Centers
  • A Senderovich
Senderovich, A.: Multi-Level Workforce Planning in Call Centers. Master's thesis, Technion (2012)
Discovering context-aware models for predicting business process performances
  • F Folino
  • M Guarascio
  • L Pontieri
  • R Meersman
  • H Panetto
  • T Dillon
  • S Rinderle-Ma
  • P Dadam
  • X Zhou
  • S Pearson