A Bayesian network approach to searching Web databases through keyword-based queries

On-line information services have become widespread in the Web nowadays. However, Web users are non-specialized and have a great variety of interests. Interfaces for Web databases must, therefore, be both simple and uniform. In this paper, we present a solution for querying Web databases using keywords only. A Bayesian network model is used to generate a set of one or more plausible structured queries derived form the initial user input. These queries can then be submitted to Web databases and the retrieved results presented as a set of ranked answers. To structure the user queries, full access to the database is not required. Instead, only a small portion of its content, extracted through a public Web interface, is used by the network model. This approach not only reduces the complexity of existing on-line interfaces, but also offers a solution to the problem of querying several distinct Web databases with a single interface.