SGER: Managing Hierarchically Organized Data in Modern Databases

层次数据库在30年前是标准的，但自那以后已经完全被关系数据库取代，尽管层次仍然是我们在其他领域组织信息的最常见方式。本研究项目的目标是根据数据库技术取得的进展重新评估层次化数据库。本文认为，关系数据库取代层次数据库并不是因为平面关系提供了更好的数据组织，而是考虑到关系技术提供的其他关键好处，即数据组织与应用程序逻辑的分离和声明性查询语言的概念。本项目的方法是从作为基本的分层数据库的轻量级目录访问协议开始，然后对其进行扩展以提供全面的数据库功能。该项目的结果将证明建立一个层次化数据库的可行性和价值，它是一种比目前流行的关系数据库更灵活和可伸缩的结构。Http://www.eecs.umich.edu/~jag/

Hierarchical databases were standard thirty years ago, but have been completely supplanted by relational databases since, in spite of hierarchy remaining the most common way in which we organize information in other spheres. The goal of this research project is to re-evaluate hierarchical databases in the light of the progress made in database technology. The thesis is that relational databases supplanted hierarchical databases not because flat relations provided a superior organization of the data, but rather on account of other crucial benefits relational technology provided, namely the separation of data organization from application logic and the notion of a declarative query language. The approach in this project is to begin with LDAP (Lightweight Directory Access Protocol) as a rudimentary hierarchical database, and to extend it to provide full-fledged database functionality. The results of this project will demonstrate the feasibility and value of building a hierarchical database as a far more flexible and scalable structure than the relational databases that are prevalent today. http://www.eecs.umich.edu/~jag/