Recommendation and diagnosis services with structure analysis of presentation documents

Abstract

The main topic addressed in this paper is how to help novice researchers compose and improve their presentation documents by means of presentation heuristics shared by the members in a laboratory. The key idea is to propose a framework of presentation structure, which represents semantic roles of and relations among presentation slides included in the documents with metadata. Following the framework, this paper introduces a machine learning technique for automatically analyzing a typical presentation structure as presentation heuristics from the repository of the documents accumulated in the laboratory. This paper also demonstrates interactive Web services that recommend the metadata to be attached to the documents newly composed, and that diagnose the presentation structure of the documents by comparison with the presentation heuristics. © 2011 Springer-Verlag.