Parallel Web search: modern application architectures

Abstract

In virtual information world, containing billions or more Web pages, search engines is still the main mechanism for accessing information. Although today the main utility of the Internet is for interactive access to documents and applications, and in almost all cases, such access is mediated by human users (typically working through Web browsers or other interactive front-end systems), the situation is constantly changing. This paper is part of a more extensive work, in which we try to find appropriate ways to deploy and use applications as remote services, over the Internet. In particular, we are interested in visualization tools and distributed programs, as front-ends for distributed databases. We would like to treat applications, to a large extent, as “black boxes”, and “wrap” them (i.e. into Java Classes and other structures), in order to enable their manipulation using increasingly available Web and Grid services infrastructures.