Facing big data variety in a model driven approach

Abstract

Despite the benefits of investing in Big Data systems are largely recognised, their adoption have been slower than expected. Actually, organisations and companies cannot migrate their systems to new a technological infrastructure without a safe integration to their legacy systems and data. For these reasons, it is required to evolve Big Data technologies with mature functions for supporting portability, interoperability and reusability. This paper illustrates a practical use case exploiting the Model-driven capabilities of the TOREADOR platform as a way to fast track the uptake of business-driven Big Data models.

Towards the 5th Industrial Revolution: A literature review and a framework for Process Optimization Based on Big Data Analytics and Semantics

“…These data integration techniques aim to combine data from various sources and provide users with a unified view of the data being analysed. The authors in [52] propose a method for dealing with large amounts of semantic data. It has four layers: the metadata layer, which describes tourism data such as geographic location and relationships between resources using different metadata rules such as MARC, DC, and GILS.…”

What Are My Users Looking for When Preparing a Big Data Campaign

“…Such procedural, platform-independent, model is finally translated into provisioning of computational resources on actual execution platforms. The semi-automatic approach of the TOREADOR platform enables seamless transformations between these models and easy deployment of a fully operational Big Data analytics tool [1].…”

“…In [7] authors highlights benefits and challenges of users' involvement in the entire development process and suggests early user involvement is directly linked with quality of requirements and end-product. By involving users'/stakeholders early in requirement elicitation by classifying and prioritizing requirements based 1 Domain analysis can elicit requirements of an existing system by studying available documentation (e.g., business plans, market studies, research studies) and previous applications to identify relevant information. on their input can be a good trade off because due to the complexity and diverse nature of a Big Data platform it is a very complex task to keep users engaged during the entire process.…”

“…Note that although no particular elicitation technique was imposed to partners, they mainly used domain analysis by capitalizing on their specific knowledge. 1 For the sake of consistency, all the partners were requested to express the elicited requirements using a uniform specification format, which includes the following information:…”

“…The TOREADOR project aims to overcome this hurdle by providing a Model-based BDA-as-a-Service framework (MBDAaaS), which will support organizations lacking in Big Data expertise in managing Big Data analytics. The TOREADOR framework includes customizable models of the entire BDA process and its artifacts [1]. As known, the first activity in the development of a new software system or technology is understanding the problem domain and what the stakeholders' needs are [2].…”