Information Extraction (IE) is the process of analyzing natural language text or speech, and collecting information about specified types of entities, relationships, or events, such as marking person, location, and organization names or determining the topic changes. As far as we know, there is no IE system for Turkish, although in recent years various Turkish processing technologies have been developed. Additionally, neither of these studies on processing Turkish text employ statistical techniques due to the problems in finding a training corpus and adapting these techniques to Turkish. As a doctoral thesis, I propose to investigate and develop a statistical information extraction system for Turkish.