Requirements-Driven Data Warehousing: A Preliminary Proof-of-Concept Study

本项目旨在对需求驱动的数据仓库构建进行初步的概念验证研究。它提出了一个新的长期愿景的第一步：一个自顶向下的数据仓库设计时以及运行时的过程，其中（大多数非技术）用户只需要专注于仔细和声明性规范的应用程序的要求在设计时，和一个工具将（半）自动生成和填充一个底层的数据仓库在运行时。 需要解决两个具体的任务：第一，设计一种概念集成建模语言，用于表示典型的集成特征，如数据起源、具体化的数据访问需求以及从概念模型到底层多维模型的映射（可以表示为数据立方体）;其次，构建一个原型工具，该工具使用以概念集成建模语言表示的模型来驱动数据仓库的设计和填充。 这项研究的预期成果将有助于加拿大商业智能行业。数据仓库是该行业的基石之一，因此研究推进数据仓库将是至关重要的旗舰加拿大软件公司，目前领先的行业。

This project aims at conducting a preliminary proof-of-concept study of a requirements-driven data warehouse construction. It proposes a first step towards a new long term vision: a top-down data warehousing design-time as well as run-time process, where (mostly non technical) users need only to focus on the careful and declarative specification of application requirements at design time, and a tool will (semi)automatically generate and populate an underlying data warehouse at run time. Two specific tasks are to be solved: first, design a conceptual integration modeling language for representing typical integration features such as data provenance, data access requirements for materialization, and mappings from the conceptual model to the underlying multidimensional model (which can be expressed as a data cube); second, build a prototype tool that uses models expressed in the conceptual integration modeling language to drive the design and the population of a data warehouse. Expected results of this research will be a contribution to the Canadian Business Intelligence industry. Data warehousing is one of the cornerstones of the industry, so research advancing data warehousing will be crucial to the flagship Canadian software companies that currently lead the industry.