A document layout system using automatic document architecture extraction

A document layout system based on automatic extraction of document architecture including logical and reference structures has been developed for reducing users' effort in document preparation, and has been implemented in a practical Japanese word processor. The extracted document architecture is used for both automatic text formatting and layout of text, figures and tables. Automatic text element recognition is performed by morphological analysis using keywords. Through intra-line (one sentence) and inter-line (relations between sentences) analysis, logical and reference structures are obtained. The automatic layout system effectively lays out the document using the extracted document architecture and knowledge about the layout.