Computer technology for preparing a natural language summary of a natural language original text based on a reading profile for an intended reader based on historical reading data. The reader profile includes reading related attribute values of the intended reader, such as levels of knowledge about various topics, a general reading level of the intended reader and the brevity/length of summaries preferred by the intended reader, the general vocabulary level of the intended reader. The intended readers reading profile will grow and change as the intended reader does additional reading to create additional new “historical” data upon which the attribute values of the reader profile are based to dynamically track attributes of the intended reader for use in further summaries of other texts.