Extracting information of future events from Arabic newspapers: an overview

Most Arabic systems developed for newspaper information extraction relate to events that occurred in the past. This paper presents a proposal for developing a web-based system that will be able to regularly collect news reports from Arabic newspaper websites, and then be able to extract information relating to future events, e.g. event type, date and location. Also, the system will be able to deposit the extracted data in an online database in order to enable users to access them. The proposed approach is based on Arabic grammar and on dependency relationships.