We demonstrate how our system can help users construct data-intensive Web sites from XML repositories and databases in a very intuitive fashion. The key idea is to show the system some example operations with data instances and let it infer the operations for the whole collection of data objects in the data repositories. This problem is challenging, because we allow data in the data repositories to be semistructured, and inferring (generalizing) the operations is a non-trivial task. We already found through experiments that even users without any knowledge on database-related concepts (such as database schemas, SQL-style queries and path expressions) can construct fairly complex Web-sites from a collection of XML documents and a relational database, by following their intuitions. We think our system is an example of how end users can be empowered by the semistructured data technologies the research community has been developing, because the Website construction process is realized by them. The demonstration is done with an implemented system of our AQUA (Amalgamation of QUerying and Authoring) project[5] . The project started with a recognition that the wide acceptance of the world wide web is not only due to its easy way to access data but also to the simple mechanism that enables end users to publish their own, personal Web contents [3]. Some of such personal Web sites are very data-intensive: For example, movie enthusiasts manage a lot of movie information, and researchers' Web sites have their publications. (They are not necessarily computer science researchers.) While they write tons of HTML documents, few of them use database management systems. We attribute it to the following two causes: (1) For end users, incorporation of databases' contents into Web sites is difficult: They can use commercially available authoring tools with drag-and-drop facilities to design their sites. However, they are typically required to write queries in SQL, which is a completely different framework from the page authoring. (2) The nature of live information from the real world and users' lack of knowledge on database principles tend to cause frequent changes on the data structures and result in semistructured data. In such situations, it is troublesome to use relational databases for data management. Based on the observation, we designed our system which (1) requires no specification of SQL-style queries in data manipulation languages, and (2) allows semistructured data repositories (such as XML document collections) to be used as underlying information sources.
展开▼
