| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| DM_AnaRibeiro_MEI_2022 | 1.9 MB | Adobe PDF |
Authors
Abstract(s)
O Extract-Transform-Load (ETL) é um componente crítico nos Sistemas de Data Warehousing (SDW) sendo responsável por extrair, transformar e carregar dados para apoiar os requisitos de tomada de decisão. Devido à complexidade da gestão dos dados, estes processos consomem grande parte dos recursos necessários na implementação dos SDW. Sendo um componente crítico que pode comprometer a adequação do sistema, se não fornecer garantias na qualidade de dados, a confiança no sistema é comprometida. Apesar da sua importância, o desenvolvimento de sistemas de ETL é essencialmente ad-hoc, o que não contribui para garantir o seguimento de práticas sólidas que garantam a coerência e coesão do desenvolvimento dos sistemas. Nos últimos anos, a Business Process Model and Notation (BPMN) tem sido proposta e utilizada para suportar os modelos conceptuais de ETL. O BPMN é uma linguagem expressiva que permite diferentes abordagens para representar os requisitos de povoamento dos processos de ETL. Neste trabalho, é explorada a utilização de BPMN para modelação conceptual de ETL, analisando as abordagens existentes e propondo um conjunto de diretrizes para utilizar o BPMN de uma forma mais consistente.
The Extract-Transform-Load (ETL) is a critical component in Data Warehousing Systems (SDW) being responsible for extracting, transforming, and loading data to support decision-making requirements. Due to the complexity of data management, these processes consume a large part of the resources needed in the implementation of SDW. Being a critical component that can compromise the suitability of the system, if it does not provide guarantees in data quality, trust in the system is compromised. Although its importance, the development of ETL systems is essentially ad-hoc, which does not contribute to guaranteeing the follow-up of solid practices that guarantee the coherence and cohesion of the development of the systems. In recent years, the Business Process Model and Notation (BPMN) has been proposed and used to support the conceptual models of ETL. BPMN is an expressive language that allows different approaches to represent the population requirements of ETL processes. In this work, the use of BPMN for conceptual modeling of ETL is explored, analyzing the existing approaches, and proposing a set of guidelines to use BPMN in a standardized way.
The Extract-Transform-Load (ETL) is a critical component in Data Warehousing Systems (SDW) being responsible for extracting, transforming, and loading data to support decision-making requirements. Due to the complexity of data management, these processes consume a large part of the resources needed in the implementation of SDW. Being a critical component that can compromise the suitability of the system, if it does not provide guarantees in data quality, trust in the system is compromised. Although its importance, the development of ETL systems is essentially ad-hoc, which does not contribute to guaranteeing the follow-up of solid practices that guarantee the coherence and cohesion of the development of the systems. In recent years, the Business Process Model and Notation (BPMN) has been proposed and used to support the conceptual models of ETL. BPMN is an expressive language that allows different approaches to represent the population requirements of ETL processes. In this work, the use of BPMN for conceptual modeling of ETL is explored, analyzing the existing approaches, and proposing a set of guidelines to use BPMN in a standardized way.
Description
Keywords
Data Warehousing Business Process Model and Notation Extract-Transform-Load modelação conceptual ETL