Abstract: A method extracts desired contents from multiple heterogeneous textual streams and provides normalized data representative of the desired contents. The method selects input streams containing text data wherein the text data of different input streams may differ in format. The method further selects a first set of parse rules corresponding to one input stream and a second set of parse rules, distinct from the first set, which correspond to a second input stream. The invention extracts desired contents from the input streams and provides normalized data which represents the desired contents. The invention selects an output interface and adapts the normalized data representing the desired contents to the output interface. The invention sends the normalized data to the output interface and the output interface is instructed to transform and format the normalized data into device specific data.