Desarrollo dirigido por modelos de interfaces específicas de dominio para el procesamiento de eventos complejos en arquitecturas orientadas a servicios

  1. Boubeta Puig, Juan
Supervised by:
  1. Inmaculada Medina Bulo Director
  2. Guadalupe Ortiz Bellot Director

Defence university: Universidad de Cádiz

Fecha de defensa: 16 July 2014

Committee:
  1. Juan Hernández Núñez Chair
  2. Juan Manuel Dodero Beardo Secretary
  3. Raúl Julián Ruggia Frick Committee member
Department:
  1. Ingeniería Informática

Type: Thesis

Teseo: 367256 DIALNET lock_openRODIN editor
Institutional repository: lockOpen access editor

Abstract

Nowadays, companies and organizations all around the world need to manage huge amounts of data from heterogeneous sources -such as own and third-party applications, web services, sensors, Internet of things platforms or social networks- every day in order to conduct the decision-making process. To be successful, a decision-making process requires, among other factors, prompt information regarding what the value of such data is for the business in question. A thorough analysis of data, as well as early action for critical or relevant situations considered as threats for an organization, will allow the organization to be positioned above its competitors. Nevertheless, it is a complex process since, among other reasons, data are heterogeneous -they do not share a common format- and they should be processed in real time. In this context, Complex Event Processing (CEP) is a technology that allows the analysis and correlation of large volumes of data with the aim of detecting complex and meaningful events, and of inferring valuable knowledge for end users. This knowledge will be really helpful in the decision-making process. To do this, so-called event patterns are used. These patterns specify which conditions must be met in order to detect such situations of interest. Despite the great advantages that CEP can bring to a business, it is a substantial challenge for users who are business experts, but do not have the necessary experience and knowledge for the use of this technology. One of the main problems these users have to face is precisely the definition of these event patterns using particular languages, those called Event Processing Languages (EPLs). Although some current software solutions provide graphical tools in order to solve this problem, none of them are user-friendly enough, since they require users to hand-write at least a portion of the code for the pattern definition. On the other hand, current information systems tend to be based on Service-Oriented Architectures (SOAs) due to the fact that this type of software architecture allows the development of highly scalable distributed systems as well as their integration with own and third-party systems. Recently, these are being combined with Event-Driven Architectures (EDAs), what is known as Event-Driven Service-Oriented Architectures (ED-SOAs or SOAs 2.0). The latter enable us to establish decoupled communications between users, applications and services. The integration of CEP with SOA 2.0 is a requirement to be able to detect real-time, relevant or critical situations in these complex and heterogeneous systems, as well as for the execution of the appropriate actions. In order to respond to these needs, the research carried out in this thesis has focused on the model-driven development of domain-specific interfaces for CEP in SOAs 2.0, with the aim of facilitating the task of defining both event patterns to be detected and alerts for real time notification for domain experts. To reach this goal, the following contributions have been supplied: a model-driven approach for CEP in SOA 2.0, a modeling language and a graphical editor for CEP domain definition, a modeling language and a reconfigurable graphical editor for event pattern definition and code generation, as well as a technological solution integrating CEP with SOA 2.0. The approach has been evaluated through the development of two case studies that have confirmed independence of the approach from the application domain. Besides, we can assert that the proposed languages are independent of event patterns and actions implementation code and that user-friendly and intuitive graphical editors hide all the implementation details from end users. This is therefore a novel approach for bringing CEP technology closer to any user, positively impacting the decision making process.