Abstract:
In this paper we give a brief overview of the state of the art in information extraction from Russian-language texts. We analyze MUC and ACE experience in event annotation. We introduce and give the definition of a model of event mention. Event mention is a syntactically connected text fragment referring to a target event of a pre-specified type. Information about the target event extracted from an event mention is used to populate an intermediate-level structure. This is assumed to be a helpful way of dealing with a great variety of textual references to the same target event. Extracting information on retirements and appointments is taken as example to discuss the challenges of fact extraction from Russian-language text. (In Russian).
Key words and phrases:automatic information extraction, factual information, test corpora, markup.