A Robust Practical Text Summarization

We present an automated method of generating human-readable summaries from text documents such as news, technical reports, government documents, and even court records. Our approach exploits an empirical observation that much of the written text display certain regularities of organization and style, which we call the Discourse Macro Structure (DMS). A summary is therefore created to reflect the con.ponents of a given DMS. In order to produce ~ roherent and readable summary we select continuoa~, well-formed passages from the source document and assemble them into a mini-document within a DMS template. In this paper we describe the SummarizerTool, a Java-implemented prototype, and its applications in various document processing tasks.