Process Instances Research Articles

Predictive process monitoring: concepts, challenges, and future research directions

Predictive Process Monitoring (PPM) extends classical process mining techniques by providing predictive models that can be applied at runtime during the execution of a business process, for example, to predict the sequence of the next event(s) in a case, its outcomes, or performance-related aspects such as the remaining processing time. These predictive models go beyond process mining’s inherent descriptive nature that is offered by typical process discovery, conformance checking, and model enhancement techniques. The growing interest in PPM is driven by its additional value proposition, i.e., delivering real-time information regarding the future execution of business process instances, thus allowing for better informed operational decision making and performance analysis. The rapid growth of PPM during the last ten years has left the field lacking a cohesive taxonomy and an explicit recognition of the prevailing challenges. This paper aims at closing this gap, by comprehensively defining PPM using a unified approach to the key terms and by discussing challenges and opportunities in the field. Specifically, we propose three overarching research challenges and nine research directions, which have been validated through a survey with PPM researchers.

Guidance for goal achievement in knowledge-intensive processes using intuitionistic fuzzy sets

Throughout the execution of a knowledge-intensive process (KiP), knowledge workers need to make critical decisions such as skipping a task or canceling a process instance. These decisions significantly impact the efficiency and effectiveness of KiP execution and should, therefore, be made in a well-informed manner. When historical data, such as event logs, is available, it can be leveraged to support knowledge workers in making these decisions. However, KiPs often lack useful historical data, as each KiP instance is unique and hardly repeatable. To address this issue, this paper proposes the novel concept of potential goal achievement, i.e., the extent to which a goal can be achieved at the end of the process, considering the collected (but incomplete) data, to support knowledge workers in efficiently executing KiPs. An approach based on Intuitionistic Fuzzy Sets (IFSs) is introduced to calculate the potential goal achievement without relying on historical data. The use of potential goal achievement in supporting knowledge workers’ decisions is demonstrated, and the effectiveness of the approach is evaluated through simulations. The results demonstrate that modeling and calculating potential goal achievement support knowledge workers in achieving goals more efficiently.

Proactive conformance checking: An approach for predicting deviations in business processes

Modern business processes are subject to an increasing number of external and internal regulations. Compliance with these regulations is crucial for the success of organizations. To ensure this compliance, process managers can identify and mitigate deviations between the predefined process behavior and the executed process instances by means of conformance checking techniques. However, these techniques are inherently reactive, meaning that they can only detect deviations after they have occurred. It would be desirable to detect and mitigate deviations before they occur, enabling managers to proactively ensure compliance of running process instances. In this paper, we propose Business Process Deviation Prediction (BPDP), a novel predictive approach that relies on a supervised machine learning model to predict which deviations can be expected in the future of running process instances. BPDP is able to predict individual deviations as well as deviation patterns. Further, it provides the user with a list of potential reasons for predicted deviations. Our evaluation shows that BPDP outperforms existing methods for deviation prediction. Following the idea of action-oriented process mining, BPDP thus enables process managers to prevent deviations in early stages of running process instances.

Spontaneous aggregation-enhanced electrochemiluminescence via galvanic strategy

Researchers unremittingly strive to develop innovative luminophores to enhance intrinsic electrochemiluminescence (ECL) performance. However, the potential to harness facile strategies, such as manipulating the physical properties of luminophores while retaining functional chemical properties to fabricate cost-effective ECL complexes, remains underexplored. Herein, we reported a novel and efficient one-step galvanic technique to actualize aggregation-enhanced ECL (AEECL) of ruthenium complexes. It marked the first instance of the galvanic process being employed to synthesize aggregate luminophores through electrostatic attraction. The ECL intensity and efficiency of the prepared ruthenium complexes with AEECL properties surpassed traditional ruthenium complexes by 8.9 and 13.6 times, respectively, outperforming most reported luminophores. Remarkably, the target luminophore exhibited high stability across varied scan rates and temperatures. Furthermore, a binder-free and carbon paper-based AEECL analytical device for lidocaine detection was fabricated, achieving a satisfactory detection limit (0.34 nM) and selectivity. The convenient modulation strategy of aggregate structure, along with the transformative leap from insufficient ECL to AEECL, bring forth a new revenue in aggregate science. This research also promises a universally applicable and versatile protocol for future biological analysis and bioimaging applications.

Task allocation for maximum cooperation in complex structured business processes

The execution of a business process usually involves the cooperation of many resources (actors) performing various tasks (activities). Generally speaking, the cooperation among actors could significantly influence the efficiency of process execution. The better cooperation among actors, the higher performance the process instance would achieve. The task allocation is, therefore, expected to ensure that the actors are compatible well with each other when performing tasks. However, the decision of task allocation is always ad-hoc or performed manually in many organizations. As for some automatic methods, they usually serve only for sequentially structured simple processes. In addition, the existing studies only concentrate on the cooperation between two actors and fail to consider the effects from the context of performing different tasks by these two actors especially in the business process with many-to-many relationships among the resources and activities. To address these issues, we handle the process with complex block structures including iteration, choice and parallelism, and propose an effective entity-wise method to measure the cooperation by combining both actors and activities together. Based on them, we present two approaches, one greedy-based and the other A*-based, to assign actors for maximum cooperation in the complex structured business process with many-to-many relationships among the resources and activities. The experimental results on both public released and synthetic datasets demonstrate the effectiveness and efficiency of our approaches.

Uncovering the origins of mental images about the concept of light: the difference between novice and expert mediation levels profile

This study investigates the mental imagery formed by two Physics undergraduates following their interaction with a mobile application analyzing emission and absorption light spectra. The tool, an instance of extracerebral cognitive processing, served to illuminate the internal cognitive mechanisms the students employed to understand novel knowledge. Guided by the Cognitive Networks Mediation Theory (CNMT), which underscores the significance of external tools in facilitating cognition, and drawing upon Bachelard's Epistemological Profile and Mortimer's Conceptual Profile, a novel 'Mediation Level' Profile is proposed. This profile represents the preferred type of mediation each student employed to elucidate their observations and understandings. The research utilized the Report Aloud Protocol to analyze students' verbal and gestural communications, revealing a complex pattern of various mental images developed through interaction with external objects via different levels of mediation. Notably, the psychophysical and hypercultural mediation levels stood out among the students. The implications of these findings for Physics education and future research directions are discussed, underscoring the utility of mobile applications as a supplementary tool in laboratory settings and the complexity of mental imagery in understanding physical phenomena.

Using virtual reality to induce multi-trial inattentional blindness despite trial-by-trial measures of awareness

Unconscious processing has been widely examined using diverse and well-controlled methodologies. However, the extent to which these findings are relevant to real-life instances of information processing without awareness is limited. Here, we present a novel inattentional blindness (IB) paradigm in virtual reality (VR). In three experiments, we managed to repeatedly induce IB while participants foveally viewed salient stimuli for prolonged durations. The effectiveness of this paradigm demonstrates the close relationship between top-down attention and subjective experience. Thus, this method provides an ecologically valid setup to examine processing without awareness.

Sentiment Analysis on Article Blog Post as an Application of NLP Chatbot

Within artificial intelligence, natural language processing is a relatively recent topic. With an estimated eight billion digital voice assistants in use due to their popularity, some of the most well-known instances of natural language processing in action include Apple Siri, Amazon Alexa, and, more recently, Google Duplex. With so much information gathered from these exchanges, further research and development on Natural Language Processing may be done, and it can be used in a number of sectors, such as business, technology, and healthcare. Sentiment analysis may be used, for example, in the healthcare industry to diagnose patients and build diagnostic models for early diagnosis of chronic disease. Sentiment analysis handles these large datasets more rapidly and efficiently with the use of Natural Language Industries. Chatbots, which have uses in customer service, healthcare, education, and workplace assistance, are becoming more important entry points to digital services and information. On the other hand, not much is known about how chatbots affect people individually, in groups, or in society as a whole. Moreover, a number of problems need to be resolved before chatbots can reach their full potential. As a result, in recent years, chatbots have become a prominent area of research. We propose a research agenda that outlines future objectives and challenges for chatbot research in order to further knowledge in this rapidly expanding field of study. This proposal synthesizes years of chatbot research debate at the CONVERSATIONS workshop series. utilizing a collaborative approach for study analysis among. Sentiment analysis is a technique or procedure used to identify and extract certain subjects from spoken and written language, such as beliefs and attitudes. Sentiment analysis, broadly speaking, is the capacity to evaluate the sentiment of a subject and categorize the general polarity of the topic phrase as positive, negative, or neutral (Kang & Park, 2014). Over the past ten years, sentiment analysis has gained significant scholarly attention.

Spatial-temporal order-disorder transition in angiogenic NOTCH signaling controls cell fate specification.

Angiogenesis is a morphogenic process resulting in the formation of new blood vessels from pre-existing ones, usually in hypoxic micro-environments. The initial steps of angiogenesis depend on robust differentiation of oligopotent endothelial cells into the Tip and Stalk phenotypic cell fates, controlled by NOTCH-dependent cell-cell communication. The dynamics of spatial patterning of this cell fate specification are only partially understood. Here, by combining a controlled experimental angiogenesis model with mathematical and computational analyses, we find that the regular spatial Tip-Stalk cell patterning can undergo an order-disorder transition at a relatively high input level of a pro-angiogenic factor VEGF. The resulting differentiation is robust but temporally unstable for most cells, with only a subset of presumptive Tip cells leading sprout extensions. We further find that sprouts form in a manner maximizing their mutual distance, consistent with a Turing-like model that may depend on local enrichment and depletion of fibronectin. Together, our data suggest that NOTCH signaling mediates a robust way of cell differentiation enabling but not instructing subsequent steps in angiogenic morphogenesis, which may require additional cues and self-organization mechanisms. This analysis can assist in further understanding of cell plasticity underlying angiogenesis and other complex morphogenic processes.

Data- & compute-efficient deviance mining via active learning and fast ensembles

Detecting deviant traces in business process logs is crucial for modern organizations, given the harmful impact of deviant behaviours (e.g., attacks or faults). However, training a Deviance Prediction Model (DPM) by solely using supervised learning methods is impractical in scenarios where only few examples are labelled. To address this challenge, we propose an Active-Learning-based approach that leverages multiple DPMs and a temporal ensembling method that can train and merge them in a few training epochs. Our method needs expert supervision only for a few unlabelled traces exhibiting high prediction uncertainty. Tests on real data (of either complete or ongoing process instances) confirm the effectiveness of the proposed approach.

Automated Software Process Discovery and Real-Time Monitoring

Process monitoring is the concept based on managing the state of a business processes executed in an enterprise software. The named status and execution time are two important parameters to infer the state of a process in real-time. An execution of the process end to end is called as process instance. This paper covers monitoring process instances in real-time based on the data metric generated from event logs, where data metrics of a Business Process Models (BPM) – a Weighted Directed Acyclic Graph (WDAG) represents process model with its heuristics such as complete execution time of process, execution time between steps/activities and the status. Firstly, the process model is discovered either by manually or automatically from the events that are recorded from the logs using unsupervised learning techniques and heuristics of different steps of a process are modeled using statistical methods. Both process model and its heuristics are deployed to help process engineers to understand the path of executions of a business process, its state and performance in real-time. Path of executions involves most successful path, longest and shortest paths, error state if the process instances abruptly ends in between, state of failure etc., On the performance front the range of time taken to complete the process, for each step, most and least time consuming, identifying the steps taking more time than usual etc., Process engineers/users are notified if abnormalities are observed. KEY WORDS Process Mining, Process Monitoring, BPM, PM4PY, Standard Deviation (SD)

Adaptive key partitioning in distributed stream processing

In stream processing systems, Key Grouping is a commonly employed partitioning scheme for distributing input tuples among parallel instances of stateful operators. With key grouping, tuples shared public keys in the stream are designated to the specific instance responsible for that key. Typically, the implementation of key grouping involves the use of a hash function. While it is convenient and deterministic, it is also known to cause load imbalance between parallel instances, especially in the presence of skewed data streams. Key-Splitting is an effective technique that distributes tasks associated with keys to downstream operators, facilitating load balancing at a relatively low cost. However, overly increasing parallel instances can lead to excessive aggregation costs, becoming a system bottleneck. In this paper, we show the high aggregation cost brought by the Key-Splitting partitioner at different levels of key separation. To address this challenge, we introduce an adaptive Key-Splitting method which controlling the degree of key separation. We propose a partitioner named FlexD, which aims to achieve dynamic adaptation of key separation limits for streaming data. The partitioner employs key grouping to distribute rare keys and dynamic expansion of processing instances to distribute hot keys. We implemented our method on Apache Storm and evaluated it by using real-world and synthetic datasets. Experimental results show that our method achieves a good balance between load balancing and aggregation cost. Moreover, it outperforms existing methods, achieving higher throughput.

Image-Mining-Based Decision Support Systems: Design Knowledge and its Evaluation in Tool Wear Analysis

Many decision processes are based on image analysis, for instance, medical diagnoses or visual monitoring of industrial processes. At the same time, advances in deep learning have significantly improved information extraction from images. While recent research strongly focuses on extracting information from single images, the potential of mining entire image collections for decision processes has been neglected so far. In this work, we develop design knowledge to use image collections for improved decision-making. We derive design requirements for image-mining-based decision support systems from literature and expert interviews. Drawing on research in image mining and decision support systems, we conceptualize design principles to address the design requirements. Subsequently, we instantiate and evaluate them in the machining industry with the help of an artifact to support tool wear analysis. The results prove the validity of our design knowledge. Our study contributes to research and practice by deriving nascent design knowledge for image-mining-based decision support systems.

Radical fluorine transfer catalysed by an engineered nonheme iron enzyme.

Nonheme iron enzymes stand out as one of the most versatile biocatalysts for molecular functionalization. They facilitate a wide array of chemical transformations within biological processes, including hydroxylation, chlorination, epimerization, desaturation, cyclization, and more. Beyond their native biological functions, these enzymes possess substantial potential as powerful biocatalytic platforms for achieving abiological metal-catalyzed reactions, owing to their functional and structural diversity and high evolvability. To this end, our group has recently engineered a series of nonheme iron enzymes to employ non-natural radical-relay mechanisms for abiological radical transformations not previously known in biology. Notably, we have demonstrated that a nonheme iron enzyme, (S)-2-hydroxypropylphosphonate epoxidase from Streptomyces viridochromogenes (SvHppE), can be repurposed into an efficient and selective biocatalyst for radical fluorine transfer reactions. This marks the first known instance of a redox enzymatic process for C(sp3)F bond formation. This chapter outlines the detailed experimental protocol for engineering SvHPPE for fluorination reactions. Furthermore, the provided protocol could serve as a general guideline that might facilitate other engineering endeavors targeting nonheme iron enzymes for novel catalytic functions.

Digital twinning for smart restoration of classic cars

Classic cars hold substantial value in the automotive industry, and the restoration process plays a pivotal role in increasing their worth. Ensuring the certification of these restoration processes is vital, as is keeping owners well-informed about the ongoing procedures taking place in their appreciated vehicles. By monitoring and controlling these restoration activities, the management of classic car workshops can effectively optimize their operations, empower owners with pertinent information, and preserve the authentic nature of these vintage automobiles.This study aims to develop a digital twin system for a classic car bodywork restoration shop workshop. The latter integrates multiple sensor technologies, including location tracking, humidity and temperature sensing, accelerometer monitoring, and smart plugs, to facilitate the identification of ongoing activities of classic car bodywork restoration process instances. By leveraging these sensors, the digital twin system may simulate and control workshop operations more effectively.We perform a systematic rapid review of related work and, based on state-of-the-art practices, we identify existing architectures and software applications used for creating digital twin systems. Then, we propose the architecture for our digital twin system, detailing its functionalities. We aim at contributing to advancing digital twin technology in classic car bodywork restoration, enhancing its authenticity, and fostering improved management practices, and overall experience for classic car owners.

Predictive processes in interpreters

Abstract This paper discusses predictive processes in simultaneous interpreting. A review of experimental findings in the interpreting process literature is carried out on prediction and anticipation – both instances of predictive processes – addressing two questions: (i) what causes prediction and anticipation?, and (ii) is there an experience-related advantage for prediction and anticipation in interpreters? A few gaps are identified in the literature, namely, the need for investigating language-specific cues to prediction and the need for investigating predictive behaviour in both source language processing and target language production. The review is followed by a discussion of future directions in addressing the identified gaps by interpreting process studies, as well as a discussion of how current and future experimental findings can inform our understanding of the processes underlying prediction during simultaneous interpreting.

Using Stocks and Flows Diagrams to Understand Business Process Behavior

Business Process Modeling has traditionally focused on the activities and logic of how work is carried out. This is depicted through modeling notations like BPMN, which illustrate the sequence of activities, performers, and possible paths for each process instance. However, when measuring the performance of an organization and its processes, the aggregation of results from individual instances is often necessary. Unfortunately, these flows are not always smooth, as they may encounter variations, delays, accumulations, and other issues that can hinder expected performance levels. Therefore, understanding the behavior of business processes over time is crucial for improvement efforts. This paper demonstrates the use of stocks and flows diagrams for modeling business processes and simulating their behavior over time. Simulations aid in identifying critical points, removing bottlenecks, and enhancing overall process performance. We begin with a brief introduction to modeling business processes using stocks and flows diagrams, followed by a real-life case study in the healthcare sector, where stocks and flows models and simulations were employed to identify and resolve a problem.

A scalable modified deep reinforcement learning algorithm for serverless IoT microservice composition infrastructure in fog layer

Nowadays many modern and Artificial Intelligence (AI) enabled Internet of Things (IoT) applications consist of chains connecting microservices distributed across the fog and cloud layers to achieve a certain functionality. These microservices should be assigned to the available processing instances placed on the things, fog and cloud layers while considering the resource utilization, Quality of Service (QoS) parameters and requirements of the user. Despite several existing optimization approaches, IoT microservice composition remains a challenge in serverless computing architecture due to the large-scale environment, its dynamic nature and the heterogeneous nodes with limited capacity. To address these challenges, a novel IoT microservice composition based on serverless architecture is proposed in the fog layer. Serverless computing is a new paradigm which offers the ability to build scalable applications without worrying about back-end management making this paradigm particularly effective in IoT environment. We specifically propose a modified Deep Reinforcement Learning (DRL)-based Microservice Chaining at Fog Layer (DRLMCF) algorithm to improve the distributed chained microservice placement in order to minimize resource utilization and delay in a fog based serverless architecture. Unlike existing methods, the proposed DRLMCF is built on realistic assumptions and does not require expert tuning or a large amount of labeled data to make optimal decisions. Furthermore, our method is not entirely dependent on a central controller and fog nodes take part in the decision making thus making the algorithm scalable in IoT environment. Compared to state-of-the art methods including load balance, shortest path and random schedule algorithms, evaluation by simulating several real-world scenarios demonstrates the proposed DRLMCF improves end to end delay up to 57.3% and the number of successfully chained microservices by up to 84%.

Multi-label learning based on instance correlation and feature redundancy

Multi-label classification is a machine-learning task that simultaneously processes instances associated with multiple labels. Label-specific feature learning selects each label's most discriminative feature subset, effectively reducing the feature dimension and improving the classification performance. However, most methods only consider label correlation, ignoring the correlation between instances and feature redundancy. To solve this problem, a multi-label classification method based on instance correlation and feature redundancy is proposed. The proposed method merges instance correlation by updating the data set and removes redundant features by calculating mutual information. By jointly considering label correlation, instance correlation, and feature redundancy, our method promotes effective multi-label feature selection. The experimental results on ten data sets demonstrate the effectiveness of the proposed method.

Expert-opinion depth-damage functions: what's the variability introduced by the survey setup?

Modelers and policy analysts need accurate depth-damage functions to forecast and mitigate potential flood damages. Often, however, these functions are either unavailable or project constraints and information barriers make it unpracticable to create them. A most frequent choice to overcome this deficiency is to repurpose existing functions developed for other contexts. Nevertheless, although expeditious, this practice introduces confidence problems to the forecasts. In these cases, expert-opinion may provide data to develop depth-damage functions that are worthy of confidence – but exactly how much? Herein we try to quantify the survey-dependent variability of expert-opinion depth-damage functions. For this it is assumed that these are but instances of a stochastic process and several synthetic surveys were simulated varying number of experts and number of calibration questions. The classical model of expert data elicitation is well suited to this experiment. Some key insights of the study were that the number of calibration questions contributes more variability than number of experts to damage functions; survey-dependent damage variability ranged, as a minimum, from ± 10 % to almost 50 %; the classical model is better for characterizing damage uncertainty than equally weighting experts’ estimates. Although further studies are needed to confirm the generalizability of these observations, these insights hopefully are useful to better appraise the implications of expert-opinion depth-damage functions for disaster management policy studies.