The Web in 2010: Challenges and Opportunities for Database Research

Abstract. The impressive advances in global networking and informa-tion technology provide great opportunities for all kinds of Web-based information services, ranging from digital libraries and information dis-covery to virtual-enterprise workflows and electronic commerce. How-ever, many of these services still exhibit rather poor quality in terms of unacceptable performance during load peaks, frequent and long out-ages, and unsatisfactory search results. For the next decade, the over-riding goal of database research should be to provide means for building zero-administration, self-tuning information services with predictable re-sponse time, virtually continuous availability, and, ultimately, “money-back ” service-quality guarantees. A particularly challenging aspect of this theme is the quality of search results in digital libraries, scientific data repositories, and on the Web. To aim for more intelligent search that can truly find needles in haystacks, classical information retrieval methods should be integrated with querying capabilities for structurally richer Web data, most notably XML data, and automatic classification methods based on standardized ontologies and statistical machine learn-ing. This paper gives an overview of promising research directions along these lines