III: Small: Information Systems Under Schema Evolution: Analyzing Change Histories and Management Tools

数据库在模式映射(如组合、可逆性)、数据交换和查询重写等方面的研究取得了重大进展，可以为数据库模式演化问题提供突破性的解决方案。但到了今天，信息系统严重缺乏应对问题所需的方法和工具，无法降低数据迁移、查询重写、应用程序重写以及架构更改造成的停机时间的成本。事实上，科学数据库(例如Ensambl)和网络信息系统(例如维基百科)的成功使这个老问题变得更加糟糕-在这些系统中，网络和科学发现过程中的应用程序和需求的快速演变因参与这些工作的用户和组织的数量和多样性而加剧。幸运的是，这些公共领域信息系统(与公司信息系统)的开放性，以及它们有趣的演变历史的丰富，使得构建一个全面的试验台来确定为这个问题提出的候选方法和工具的优点、局限性和潜力是可能的。因此，这个项目正在建设：(I)一个包含关键信息系统进化历史的开放源码精选储存库，(Ii)在其中测试的一套全面工具的基准，以及(Iii)收集和分析进化历史的工具。这些工具然后被用来(A)比较和评估现有的方法、方法和工具，(B)吸引研究人员评估和改进他们的技术，并将他们的测试案例添加到基准中。可以预期对模式映射研究和应用产生变革性的影响，因为理论解决方案现在在现实生活案例研究中得到验证和改进。这些反过来有望极大地改变和改进科学数据库和网络信息系统。有关更多信息，请参阅项目Web page:http://www.cs.ucla.edu/~zaniolo/nsf0917333.html

The significant progress made by database research on schema mapping (e.g., omposition, invertibility), data exchange, and query rewriting, can provide breakthrough solutions for the Database Schema Evolution problem. But as of today, information systems are sorely lacking the methods and tools needed to cope with theproblem, and to reduce the cost of data migration, rework of queries, application rewriting, and downtime created by schema changes. In fact, this old problem has been made worse by the success of scientific databases (e.g., Ensembl) and web informationsystems (e.g., Wikipedia)---where the fast evolution of applications and requirements characterizing the web and the scientific discovery process is exacerbated by the number and diversity of users and organizations cooperating on these endeavors.. Fortunately, the openness ofthese public-domain information systems (vs. corporate ones), and the abundance of their interesting evolution histories make it possible tobuilt a comprehensive testbed to determine the strengths, limitations, and potentials of candidate methods and tools proposed for the problem. Thus, this project is building: (i) an open-source curated repository containing evolution histories from key information systems, (ii) benchmarks for a comprehensive set of tools tested therein, and (iii) instruments to collect and analyze evolution histories.These are then used to (a) compare and evaluate existing approaches, methods and tools, and (b) entice researchers to evaluate and improve their techniques and add their test cases to the benchmark.A transformative impact can be expected upon schema mappingresearch and applications, inasmuch as theoretical solutions are now validated and improved on real-life case-studies. These in turn are expected to transform and improve significantly scientific databases and web information systems. For furtherinformation see the project web page:http://www.cs.ucla.edu/~zaniolo/nsf0917333.html