An intelligent system for retrieving economic information from corporate websites

The prompt availability of up-to-date economic in-dicators is crucial to monitor the economy and to steer the design of policies for promoting business innovation and raising firm competitiveness. Economic indicators usually suffer important lags since they are commonly obtained from official databases or from interviews to a sample of agents; thus limiting the representativeness and usefulness of the information. In a context in which the presence of companies in the World Wide Web is al-most an obligation to succeed, corporate websites are connected, in some way, to the firm economic activ-ity. On the basis of this relation, this paper proposes an intelligent system that analyzes corporate websites to produce web indicators related to the economic ac-tivity of the firms. This system has been successfully implemented and applied to infer company size char-acteristics from data gathered from corporate websites. Our results show that relatively large companies pro-vide web content in a foreign language and use propri-etary web servers. 1