Advanced Interfaces to Scientific Databases

在过去的十年中，实验科学家个人生成和存储数据的能力有了显著的提高，而数据管理工具却没有相应的发展。ZOO是一个通用的桌面实验管理环境，可以定制并放置在许多科学家的桌子上来管理他们的实验研究。对Zoo第一个原型的早期体验表明，与一般数据库系统一样，最大的挑战之一在于系统的用户界面。这个项目的目标是研究与高级数据库用户界面相关的几个难题，这些问题对实验管理显得特别重要。这些问题包括与模式可视化、实验流建模、可视化查询语言、动态查询、近似查询答案、增量查询等相关的问题。这项工作在开发通用数据库用户界面技术、将结果技术合并到ZOO中以及在实际实验环境中安装工具的定制增强以进行评估和反馈之间的紧密循环中进行。这项研究的预期结果将推动数据库用户界面的发展，并将被纳入到ZOO工具中。有了这种方式，ZOO将达到必要的成熟度，可以放置在实际的实验生产环境中，使科学家能够高效地管理他们的实验研究。

Over the last decade, a dramatic increase has been observed in the ability of individual experimental scientists to generate and store data, which has not been matched by an equivalent development of adequate data management tools. ZOO is a generic Desktop Experiment Management Environment that can be customized and placed on the desks of many scientists to manage their experimental studies. Early experiences with a first prototype of Zoo have indicated that, as with general database systems, one of the biggest challenges lies with the system's user interface. The goal of this project is to investigate several hard problems related to advanced database user interfaces that appear particularly important for experiment management. These include problems related to schema visualization, experiment flow modeling, visual query languages, dynamic queries, approximate query answers, incremental queries, and others. This work proceeds in a tight loop between developing generic database user interface technology, incorporating the resulting techniques into ZOO, and installing customized enhancements of the tool in real-life experimentation environments for evaluation and feedback. The expected results of this research will advance the state of the art in database user interfaces and will be incorporated into the ZOO tools. Empowered with this fashion, ZOO will reach the necessary maturity to be placed in actual experiment production environments and enable scientists to manage their experimental studies efficiently and effectively.