Data-Driven Analysis of Batch Processing Inefficiencies in Business Processes

Predictive business process monitoring aims at providing the predictions about running instances by analyzing logs of completed cases of a business process. Recently, a lot of research focuses on increasing productivity and efficiency in a business process by forecasting potential problems during its executions. However, most of the studies lack suggesting concrete actions to improve the process. They leave it up to the subjective judgment of a user. In this paper, we propose a novel method to connect the results from predictive business process monitoring to actual business process improvements. More in detail, we optimize the resource allocation in a non-clairvoyant online environment, where we have limited information required for scheduling, by exploiting the predictions. The proposed method integrates offline prediction model construction that predicts the processing time and the next activity of an ongoing instance using LSTM with online resource allocation that is extended from the minimum cost and maximum flow algorithm. To validate the proposed method, we performed experiments using an artificial event log and a real-life event log from a global financial organization.

Everything at the proper time: Repairing identical timestamp errors in event logs with Generative Adversarial Networks

Evaluating and predicting overall process risk using event logs

Business processes are implemented in an organization. When a business process is run, it generates event log. One type of event log is double timestamp event log. Double timestamp has the start and complete time of each activity executed in the business process and has a close relationship with temporal pattern. In this paper, seven types of temporal pattern between activities were presented as extended version of relations used in the double timestamp event log. Since the event log was not always executed in sequential way, therefore using temporal pattern, event log was divided into several small groups to mine the business process both sequential and parallel. Both temporal pattern and Time-based α+ Miner algorithm were used to mine process model, determined sequential and parallel relations and then evaluated the process model using fitness value. This paper was focused on the advantages of temporal pattern implemented in Time-based α+ Miner algorithm to mine business process. The results also clearly stated that the proposed method could present better result rather than that of original α+ Miner algorithm.

비즈니스 프로세스 및 워크플로우 기술의 국제표준화기구인 WfMCl)에서는 최근 비즈니스 프로세스 인텔리전스 마이닝 분야에 대한 산업체의 관심이 증가함에 따라 프로세스 실행이벤트로그 표준포맷인 비즈니스 프로세스 분석로그 포맷, BPAF2) 1.0을 공식적으로 발표한 바 있다. 즉, 비즈니스 프로세스 인텔리전스 마이닝 기술은 비즈니스 프로세스 모델의 실행이벤트로그로부터 제어흐름, 데이터흐름, 역할흐름, 수행자흐름 등의 흐름중심의 인텔리전스와 최근에 관심이 집중되는 프로세스기반 소셜네트워크, 소속성네트워크 등의 관계중심의 인텔리전스를 마이닝하는 일련의 알고리즘들과 분석기법들로 구성되는데 현재의 표준포맷인 BPAF 1.0은 비즈니스 프로세스의 제어흐름 인텔리전스 마이닝에 초점 맞추고 있어 최근에 관심이 집중되는 관계중심의 인텔리전스 마이닝을 지원할 수가 없다. 따라서, 본 표준화 논문에서는 제어흐름 인텔리전스 이외에 데이터흐름, 역할흐름, 수행자흐름의 흐름 중심 인텔리전스 뿐만 아니라 프로세스기반 소셜네트워크, 소속성 네트워크의 관계중심 인텔리전스의 마이닝을 지원할 수 있도록 기존의 BPAF 1.0 표준포맷을 확장한 BPAF 2.0 표준포맷을 제안한다. 특히, 본 논문에서 제안하는 BPAF 2.0은 한국정보통신기술협회 표준총회의 e 비즈니스 프로젝트 그룹을 통한 국내 표준안의 기반기술이 될 뿐 만 아니라 BPAF 1.0을 제정한 WfMC 국제표준화기구의 국제 표준안의 확장에 기여할 것이라고 판단한다. WfMC, which is one of the international standardization organizations leading the business process and workflow technologies, has been officially released the BPAF1.0 that is a standard format to record process instances' event logs according as the business process intelligence mining technologies have recently issued in the business process and workflow literature. The business process mining technologies consist of two groups of algorithms and their analysis techniques; one is to rediscover flow-oriented process-intelligence, such as control-flow, data-flow, role-flow, and actor-flow intelligence, from process instances' event logs, and the other has something to do with rediscovering relation-oriented process-intelligence like process-driven social networks and process-driven affiliation networks from the event logs. The current standardized format of BPAF1.0 aims at only supporting the control-flow oriented process-intelligence mining techniques, and so it is unable to properly support the relation-oriented process-intelligence mining techniques. Therefore, this paper tries to extend the BPAF1.0 so as to reasonably support the relation-oriented process-intelligence mining techniques, and the extended BPAF is termed BPAF2.0. Particularly, we have a plan to standardize the extended BPAF2.0 as not only the national standard specifications through the e-Business project group of TTA, but also the international standard specifications of WfMC.

Process mining enables extracting insights into human resources working in business processes and supports employee management and process improvement. Often, resources from the same organizational group exhibit similar characteristics in process execution, e.g., executing the same set of process activities or participating in the same types of cases. This is a natural consequence of division of labor in organizations. These characteristics can be organized along various process dimensions, e.g., case, activity, and time, which ideally are all considered in the application of resource-oriented process mining, especially analytics of resource groups and their behavior. In this paper, we use the concept of execution context to classify cases, activities, and times to enable a precise characterization of resource groups. We propose an approach to automatically learning execution contexts from process execution data recorded in event logs, incorporating domain knowledge and discriminative information embedded in data. Evaluation using real-life event log data demonstrates the usefulness of our approach.KeywordsExecution contextResource groupEvent logProcess miningWorkforce analytics

In a real-world environment, business processes are complex and challenging to monitor for potential irregularities. Service-oriented organisations rely primarily on the effectiveness of their processes to maintain the quality of their offered services and prefer to be notified in advance if an ongoing business process is not executing as expected. Process mining techniques analyse business processes, but heterogeneity & variability in the real-world event logs make it challenging and time-consuming to mine using standard methods. Trace clustering divides event logs into sub-logs with similar properties based on the case and event attributes. However, given the size of the event logs collected from heterogeneous information systems, it is difficult to identify meaningful clusters in actual business process logs. This paper proposes a hybrid technique to decompose large event logs into several smaller sub-logs, making it easier and more time efficient to analyse. Each sub-log is independently investigated to gain valuable insights into the process's underlying behaviour. Non-traditional process mining techniques are applied to identify possible behavioural correlations and discrepancies between simulated and observed process execution by extracting features from the log segment. We used a real-world case study to depict our framework's usefulness in prediction accuracy and timeliness. Furthermore, we compared our findings to those of previously proposed methodologies from the literature. We demonstrated that the framework improved the fitness quality of the resultant business process models.

Process mining is a discipline that uses techniques to extract knowledge from event logs recorded by information systems in most companies these days. Among main perspectives of process mining, organizational and time perspectives focus on information about resources stored on the event logs and timing and frequency of the events, respectively. In this paper we introduce a method that combines organizational and time perspectives of process mining with a decision support tool called decision trees. The method takes the information of historical process data by means of an event log, generates a decision tree, annotates the decision tree with processing times, and recommends the best performer for a given running instance of the process. We finally illustrate the method through several experiments using a developed plug-in for the process mining framework ProM, first using synthetic data and then using a real-life event log.

Real-life business process event logs may suffer from significant data quality problems negatively influencing process mining analysis. Over time, a range of approaches has been developed to detect and repair these quality problems. Validation of these approaches tends to be challenging due to the lack of a ground truth. Moreover, the identification and definition of event log quality problems have been tackled mainly through a pattern-based approach, with systematic and extensible methods currently lacking. In this article, we present FLAWD , a formal language for describing event log data quality issues that enables solutions addressing the shortcomings of process mining data quality research identified above. FLAWD can be used to formally describe and possibly reason over event log data quality errors, as well as to guide the development of tools for controlled and sophisticated “polluting” of event logs through which benchmark datasets may be systematically created. We present the abstract syntax grammar of FLAWD and an open-source software tool based on it that allows for the insertion of all so-called event log imperfection patterns in a stochastic manner. We show how FLAWD has been used in our research to generate benchmark datasets and how it can be used to formally describe and replicate a range of errors found in real-life event logs.

Many present-day companies carry out a huge amount of daily operations through the use of their information systems without ever having done their own enterprise modeling. Business process mining is a well-proven solution which is used to discover the underlying business process models that are supported by existing information systems. Business process discovery techniques employ event logs as input, which are recorded by process-aware information systems. However, a wide variety of traditional information systems do not have any in-built mechanisms with which to collect events (representing the execution of business activities). Various mechanisms with which to collect events from non-process-aware information systems have been proposed in order to enable the application of process mining techniques to traditional information systems. Unfortunately, since business processes supported by traditional information systems are implicitly defined, correlating events into the appropriate process instance is not trivial. This challenge is known as the event correlation problem. This paper presents an adaptation of an existing event correlation algorithm and incorporates it into a technique in order to collect event logs from the execution of traditional information systems. The technique first instruments the source code to collect events together with some candidate correlation attributes. Based on several well-known design patterns, the technique provides a set of guidelines to support experts when instrumenting the source code. The event correlation algorithm is subsequently applied to the data set of events to discover the best correlation conditions, which are then used to create event logs. The technique has been semi-automated to facilitate its validation through an industrial case study involving a writer management system and a healthcare evaluation system. The study demonstrates that the technique is able to discover an appropriate correlation set and obtain well-formed event logs, thus enabling business process mining techniques to be applied to traditional information systems.

Most actual information systems save transactions in event logs. These logs are not often used to improve the quality of the process model. The area of Process Mining provides techniques and tools which allow the extraction of persistent information about the business process model from event logs in order to incorporate them in a decisional system, and optimise the future executions of the business processes. In this paper, we present an approach using (in addition of event logs) some properties of the tasks composing the business processes like minimum and maximum durations. These properties are used to improve the quality of the underlying processes by a best detection of parallelism, loops and duplicated tasks.

In previous chapters, we have discussed the use of the multilevel business artifact (MBA) for multilevel business process modeling in order to better represent the interactions between the different hierarchy levels within a company. As the preparation for a proof-of-concept implementation, in Chapter 6, we have introduced an XML-based logical representation for MBAs. Relying on this XML representation, in this chapter, we present the implementation concept for a business process management system that enables the MBA-based (semi-)automated execution of multilevel business processes. A business process management system manages process-related data and allows for the event-driven execution of business processes; a business process management system also records event data. The recording of event data in event logs is a pre-requisite for business process intelligence which allows for performance analysis and the subsequent improvement of business processes based on the analysis results.

Conformance checking is a method to compare the actually executed business process recorded as an event log with the business process described as a business process model and to identify differences. For human or technical reasons, event logs that contain noise and are of low quality may be recorded. Therefore, methods have been proposed to repair low-quality event logs, but they require the setting of a threshold, and it is difficult to set a suitable threshold. In this paper, we investigate the effect of low-quality event log repair methods on conformance checking. Through experiments, it was shown that the appropriate threshold depends on the type of event log and the amount of noise.

The goal of process mining is to discover the process behavior from the runtime information of process executions. Having labeled process logs is crucial for process mining research. However, real life event logs at process-aware information systems are mostly partially assigned to case identifiers, known as unlabeled event log problem. As a remedy to labeled data need in process mining research, we propose an approach to generate synthetic event logs according to the provided process profile, which outlines the activity vocabulary and structure of the corresponding business process. We evaluate the performance of our prototypical implementation in term of compatible log generation under varying parameter setting complexities.

The current business environment has no room for inefficiency as it can cause companies to lose out to their competitors, to lose customer trust, and to experience cost overruns. Business processes within the company continue to grow and cause them to run more complex. The large scale and complexity of business processes pose a challenge in improving the quality of process model because the effectiveness of time and the efficiency of existing resources are the biggest challenges. In the context of optimizing business processes with a process mining approach, most current process models are optimized with a trace clustering approach to explore the model and to perform analysis on the resulting process model. Meanwhile, in the event log data, not only the activities but also the other resources, such as records of employee or staff working time, process service time, and processing costs, are recorded. This article proposes a mechanism alternative to optimize business processes by exploring the resources that occur in the process. The mechanism is carried out in three stages. The first stage is optimizing the job shop scheduling method from the generated event log. Scheduling the time becomes a problem in the job shop. Utilizing the right time can increase the effectiveness of performance in order to reduce costs. Scheduling can be defined as the allocation of multiple jobs in a series of machines, in which each machine only does one job at a time. In general, scheduling becomes a problem when sequencing the operations and allocating them into specific time slots without prolonging the technical and capacity constraints. The second stage is generating the resource value that is recorded in the event log from the results of analysis of the previous stage, namely, job shop scheduling. The resource values are multivariate and then clustered to determine homogeneous clusters. The last stage is optimizing the nonlinear multipolynomials in the homogeneous cluster formed by using the Hessian solution. The results obtained are analyzed to get recommendations on business processes that are appropriate for the company’s needs. The impact of long waiting times will increase service costs, but by improving workload, costs can be reduced. The process model and the value of service costs resulting from the mechanism in the research can be a reference for process owners in evaluating and improving ongoing processes.