Informatică economică (Jan 2010)
The Development and the Evaluation of a System for Extracting Events from Web Pages
Abstract
The centralization of a particular event is primarily useful for running news services. These services should provide updated information, if possible even in real time, on a specific type of event. These events and their extraction involved the automatic analysis of linguistic structure documents to determine the possible sequences in which these events occur in documents. This analysis will provide structured and semi-structured documents in which the unit events can be extracted automatically. In order to measure the quality of a system, a methodology will be introduced, which describes the stages and how the decomposition of a system for extracting events in components, quality attributes and properties will be defined for these components, and finally will be introduced metrics for evaluation.
