Construction of Dependency Structure Matrix (DSM) for e-books using ontology based text mining

Abstract

Dependency Structure Matrix (DSM) is a square matrix, used to map the dependencies between each entity/element. Such dependencies are examined to make out impacts. In general, college libraries maintain e-books for various departments, as students, research scholars in these days have started to refer e-books in most of the times. Since we don't have hard copies for some of the books, we go for searching e-books. This paper primarily focuses on finding related/linked/correlated/interrelated e-books that for a targeted e-book based on table of contents of each book. Related books will hold desired content that we require. Extract content page to ascertain the classes of e-books, ontology that provides inferences, regarding them. Subsequently, uses this ontology to discover the dependencies between each book and construct DSM which is used to represent related books as a result of comparing table of contents of each book. Subsequent to construction of DSM, it is converted into graph structure along with dependency level for easy analysis and find out. During arrival of fresh e-book to library, DSM is automatically updated by mining PDF and the graph is also updated along with their dependency level.