An Algorithm for One-page Summarization of a Long Text Based on Thematic Hierarchy Detection

This paper presents an algorithm for text summarization using the thematic hierarchy of a text. The algorithm is intended to generate a one-page summary for the user, thereby enabling the user to skim large volumes of an electronic book on a computer display. The algorithm first detects the thematic hierarchy of a source text with lexical cohesion measured by term repetitions. Then, it identifies boundary sentences at which a topic of appropriate grading probably starts. Finally, it generates a structured summary indicating the outline of the thematic hierarchy. This paper mainly describes and evaluates the part for boundary sentence identification in the algorithm, and then briefly discusses the readability of one-page summaries.