A knowledge-base document analysis system

This paper describes a document analysis system capable to extract layout structure and determine significant logical objects on printed documents without any text recognition. The system adapts algorithms based on different knowledge sources that ponder geometric aspects of information arrangements and geometric characteristics of information parts. The system utilizes these knowledge sources to construct a knowledge base for document structure analysis and logical labeling, in order to provide Facilities for information retrieval of documents. Results from experiments demonstrate that the proposed system is effective for practical applications.