Watermarking generative information systems for duplicate traceability

This document outlines a watermarking scheme applicable to generative information systems in general, and dynamic webpage content delivered by a HTTP server in particular. The approach focuses on tracability of potential duplicates of previously served informational content. The suggested procedures are non-intrusive, resilient, blind, robust, reversible and suitable for large amounts of textual content. In order to do so, non-or semi-printable characters are inserted at syntactically and semantically valid places within the payload data.