Design and development of extended text feature classification methodology for effective information extraction