摘要 |
This invention relates to a method, system and computer program product for annotating a document, said document being divided by intervals, said method comprising: identifying, at each interval in the document, one or more keywords in the text of the document; determining, at each interval, the number of context classes associated with each keyword; identifying at least one annotation word and at an interval in the text of the document, said annotation word having two or more annotations associated with two or more context classes; determining the annotation based on the context class with highest count at the annotation word interval; and annotating the document with the determined annotation.
|