Presents an information extraction mechanism that can process many kinds of inputs, realize the type of text, and understand the percentage of the keywords that has to be stored. This mechanism then supports information extraction and information categorization mechanisms.