Facing the increasing global competition, modern business organizations have to respond quickly and correctly to the constant changing environment to ensure their competitive advantages. This goal has led to a recent surge of work on Business Process Reengineering (BPR) and Workflow Management. While most work in these areas assume that process definitions are known in a priori, it is widely recognized that defining a process type which totally represents all properties of the underlying business process is a difficult job. This job is currently practiced in a very ad-hoc fashion. In this paper, we postulate an algorithm to discover the process definition from analyzing the existing process instances. We compare our algorithm with other existing algorithms proposed in the literature in terms of time complexity and apply these algorithms through synthetic data sets to measure the qualities of output results. It has been found that our algorithm is able to return the process definitions closer to the real ones in a faster manner.

The volume of data related to business process execution is increasing significantly in the enterprise. Many of data sources include events related to the execution of the same processes in various systems or applications. Event correlation is the task of analyzing a repository of event logs in order to find out the set of events that belong to the same business process execution instance. This is a key step in the discovery of business processes from event execution logs. Event correlation is a computationally-intensive task in the sense that it requires a deep analysis of very large and growing repositories of event logs, and exploration of various possible relationships among the events. In this dissertation, we present a scalable data analysis technique to support efficient event correlation for mining business processes. We propose a two-stages approach to compute correlation conditions and their entailed process instances from event logs using MapReduce framework. The experimental results show that the algorithm scales well to large datasets.

This Master's describes knownledge discovery from process logs by using process mining algorithms. Chosen algorithms are described in detail. These aim to create process model based on event log analysis. The goal is to design such components, which would be able to import the process and run the simulations. Results from components can be used for short term planning.

The managing of complex business processes, which are changed due to globalization, calls for the development of powerful information systems that offer generic process modeling and process execution capabilities. Even though contemporary information systems are more and more utilized in enterprises, their actual impact in automatizing complex business process is still limited by the difficulties encountered in design phase. Actually this design phase is time consuming, often subjective and incomplete. In the scope of this study, a reverse approach is followed. Instead of starting with process design, the method of discovering interesting patterns from the navigation traces is taken as basis and a new data analysis methodology named &ldquo / From-to Chart Based Process Discovery&rdquo / is proposed. In this hybrid methodology &ldquo / from-to chart&rdquo / , which is fundamentally dedicated to material handling issues on production floor, is used as the front-end to monitor the transitions among activities of a realistic event log and convert these raw relations into optimum activity sequence. Then a revised version of process mining, which is the back-end of this methodology, upgrades optimum activity sequence into process model.

Submitted by Luciana Ferreira (lucgeral@gmail.com) on 2017-07-24T11:15:23Z No. of bitstreams: 2 Dissertação - Caio Appelt Rezende - 2017.pdf: 1362955 bytes, checksum: b87d99aa29bcdf87d151fb1d32bb57ee (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-07-24T11:15:36Z (GMT) No. of bitstreams: 2 Dissertação - Caio Appelt Rezende - 2017.pdf: 1362955 bytes, checksum: b87d99aa29bcdf87d151fb1d32bb57ee (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-07-24T11:15:36Z (GMT). No. of bitstreams: 2 Dissertação - Caio Appelt Rezende - 2017.pdf: 1362955 bytes, checksum: b87d99aa29bcdf87d151fb1d32bb57ee (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-05-03 / Process Mining is a recent area of research and is composed of techniques that allow the analysis and extraction of knowledge from the logs of the business processes obtained from Management Information Systems (MIS). The analyzes can be classified into three types: Process Discovery, Conformance Check and Process Improvement. With the current growth not only of quantity, but also of the types of algorithms that seek to fulfill the objectives of Process Mining, a classification that takes into account the performance of the algorithm in the various real situations of its application becomes important. The Evaluation and Comparison of the algorithms from the repository data could be done through the application of Quality Metrics or Machine Learning Techniques. This work presents a proposal of a set of Quality Metrics to allow the classification, evaluation and comparison of Process Discovery algorithms. The proposal is based on the review of algorithms and their families; the possible performance characteristics, that can be applied to any type of algorithm being tested; and in simulations of business process patterns. The results obtained by the work are promising in the sense of creating the conceptual basis and a methodology for future research to allow the construction of a framework for Evaluation and Comparison of new algorithms. / A Mineração de Processos (Process Mining) é uma área de pesquisa recente e é composta por técnicas que permitem a análise e a extração de conhecimento a partir dos registros de eventos (logs) dos processos de negócios obtidos de Sistemas de Informação Gerenciais (SIG). As análises podem ser classificadas em três tipos: Descoberta de Processos, Checagem da Conformidade e Melhoria de Processos. Com o atual crescimento não apenas da quantidade, mas também dos tipos de algoritmos que procuram cumprir os objetivos da Mineração de Processos, uma classificação que leve em consideração a performance do algoritmo nas diversas situações reais de sua aplicação se torna importante. A Avaliação e a Comparação dos algoritmos a partir dos dados do repositório poderiam ser feitas através da aplicação de Métricas de Qualidade ou Técnicas de Aprendizado de Máquina. Este trabalho apresenta uma proposta de um conjunto de Métricas de Qualidade que tem como objetivo permitir a classificação, avaliação e comparação de algoritmos de Descoberta de Processos. A proposta foi construída com base na revisão dos algoritmos e suas famílias; no levantamento das possíveis características de performance, que podem ser aplicadas a qualquer tipo de algoritmo sendo testado; e em simulações de registros de eventos de padrões de processos de negócio. Os resultados obtidos pelo trabalho são promissores no sentido de criar a base conceitual e uma metodologia para que futuras pesquisas permitam a construção de um arcabouço (framework) de Avaliação e Comparação de novos algoritmos.

Business process analysis and integration are key endeavours for today's enterprises. Recently, Web services have been widely adopted for the implementation and integration of business processes within and across enterprises. In this dissertation, we investigate the problem of enabling the analysis of service interactions, in today's enterprises, in the context of business process executions, and that of service integration. Our study shows that only fraction of interactions in the enterprise are supported by process-aware systems. However, enabling above-mentioned analyses requires: (i) a model of the underlying business process to be used as a reference for the analysis, and (ii) the ability to correlate events generated during service interactions into process instances. We refer to a process model and the corresponding process instances as a "process view". We propose the concept of process space to refer to all process related information sources in the enterprise, over which various process views are defined. We propose the design and development of a system called "process space discovery system" (PSDS) for discovering process views in a process space. We introduce novel approaches for the correlation of events into process instances, focusing on the public processes of Web services (business protocols), and also for the discovery of the business protocol models from the process instances of a process view. Analysis of service integration approaches shows that while standardisation in Web services simplifies the integration in the communication level, at the higher levels of abstractions (e.g., services interfaces and protocol models) services are still open to heterogeneities. We characterise the mismatches between service interfaces and protocol specifications and introduce "mismatch patterns" to represent them. A mismatch pattern also includes an adapter template that aims at the resolution of the captured mismatch. We also propose semi-automated approaches for identifying the mismatches between interface and protocol specifications of two services. The proposed approaches have been implemented in prototype tools, and experimentally validated on synthetic and real-world datasets. The discovered process views, using PSDS, can be used to perform various analyses in an enterprise, and the proposed adaptation approach facilitates the adoption of Web services in business process integration.

Business process analysis and integration are key endeavours for today's enterprises. Recently, Web services have been widely adopted for the implementation and integration of business processes within and across enterprises. In this dissertation, we investigate the problem of enabling the analysis of service interactions, in today's enterprises, in the context of business process executions, and that of service integration. Our study shows that only fraction of interactions in the enterprise are supported by process-aware systems. However, enabling above-mentioned analyses requires: (i) a model of the underlying business process to be used as a reference for the analysis, and (ii) the ability to correlate events generated during service interactions into process instances. We refer to a process model and the corresponding process instances as a "process view". We propose the concept of process space to refer to all process related information sources in the enterprise, over which various process views are defined. We propose the design and development of a system called "process space discovery system" (PSDS) for discovering process views in a process space. We introduce novel approaches for the correlation of events into process instances, focusing on the public processes of Web services (business protocols), and also for the discovery of the business protocol models from the process instances of a process view. Analysis of service integration approaches shows that while standardisation in Web services simplifies the integration in the communication level, at the higher levels of abstractions (e.g., services interfaces and protocol models) services are still open to heterogeneities. We characterise the mismatches between service interfaces and protocol specifications and introduce "mismatch patterns" to represent them. A mismatch pattern also includes an adapter template that aims at the resolution of the captured mismatch. We also propose semi-automated approaches for identifying the mismatches between interface and protocol specifications of two services. The proposed approaches have been implemented in prototype tools, and experimentally validated on synthetic and real-world datasets. The discovered process views, using PSDS, can be used to perform various analyses in an enterprise, and the proposed adaptation approach facilitates the adoption of Web services in business process integration.

Le volume des données relatives à l'exécution des processus métiers augmente de manière significative dans l'entreprise. Beaucoup de sources de données comprennent les événements liés à l'exécution des mêmes processus dans différents systèmes ou applications. La corrélation des événements est la tâche de l'analyse d'un référentiel de journaux d'événements afin de trouver l'ensemble des événements qui appartiennent à la même trace d'exécution du processus métier. Il s'agit d'une étape clé dans la découverte des processus à partir de journaux d'événements d'exécution. La corrélation des événements est une tâche de calcul intensif dans le sens où elle nécessite une analyse approfondie des relations entre les événements dans des dépôts très grande et qui évolue de plus en plus, et l'exploration de différentes relations possibles entre ces événements. Dans cette thèse, nous présentons une technique d'analyse de données évolutives pour soutenir d'une manière efficace la corrélation des événements pour les fouilles des processus métiers. Nous proposons une approche en deux étapes pour calculer les conditions de corrélation et héritier entraîné des instances de processus de journaux d'événements en utilisant la plateforme MapReduce. Les résultats expérimentaux montrent que l'algorithme s'adapte parfaitement à de grands ensembles de données. / The volume of data related to business process execution is increasing significantly in the enterprise. Many of data sources include events related to the execution of the same processes in various systems or applications. Event correlation is the task of analyzing a repository of event logs in order to find out the set of events that belong to the same business process execution instance. This is a key step in the discovery of business processes from event execution logs. Event correlation is a computationally-intensive task in the sense that it requires a deep analysis of very large and growing repositories of event logs, and exploration of various possible relationships among the events. In this dissertation, we present a scalable data analysis technique to support efficient event correlation for mining business processes. We propose a two-stages approach to compute correlation conditions and their entailed process instances from event logs using MapReduce framework. The experimental results show that the algorithm scales well to large datasets.

[pt] Descoberta de Processo significa determinar um modelo de processo a partir de um registro histórico de eventos de um processo de negócios. Muitos algoritmos de descoberta de processos tentam sintetizar uma rede de Petri que representa o registro localizando locais e arcos que relacionam as classes de eventos. Bergenthum et al (2007) e van der Werf et al. (2008) propõem formulações para este problema descobrir um place de cada vez, em que cada solução básica do conjunto de desigualdades representa um lugar candidato. Propomos uma formulação global de programação inteira que, dado um registro histórico, determina todos os places e arcos que definem uma rede de Petri de uma só vez. Este modelo é uma alternativa a seleção de locais, mas tem um problema de eficiência devido à grande quantidade de variáveis inteiras usadas. Também propomos um método de decomposição para o modelo ILP global para tratar cada place e suas restrições associadas como um subproblema separado. Conseguimos então executar o algoritmo em instâncias sintéticas grandes, o que é inédito para esta classe de mineradores de processo. / [en] Process Discovery amounts to determine a process model from an event log of a business process. Many process discovery algorithms try to synthesize a Petri net representing the log by finding places and arcs that relate the event classes. Bergenthum et al. (2007) and van der Werf et al. (2008) propose formulations for this problem discover one place at a time, in which each basic solution of the set of inequalities represents a candidate place. We propose a global integer programming formulation that, given a log, determines all places and arcs defining a Petri net. This model simplifies the selection of places but has an efficiency problem due to a large number of integer variables used. We also propose a decomposition method for the global ILP model to treat each place and their associated constraints as a separate sub-problem. We can run the algorithm on large synthetic instances, which is unprecedented for this kind of process miner.

No description available.