A Fragment Matching Evaluation Method Based on Character Structure

Abstract

Information recovery is an important research topic in the field of information security, and assessing the degree to which fragments match is a key step in recovery algorithm. In this study, based on a character's structure in real document fragments, a novel measurment function for fragment matching is presented that can evaluate the degree of correlation between the fragments. In addition, the function is used as the judgment basis in the fragment search algorithm to determine whether a found match is the global optimum. The experimental results show that the proposed method can effectively inhibit information damage and noise interference that exist in real document fragments, and that it accurately reflects the degree to which the fragments match. The information in document can be effectively restored by this method.