The Construction And Application Of Concept-Attribute Knowledge Network Based On Digital Books

With the rapid development of information technology and the popularity of Internet technology, the amount of data accumulated is growing rapidly. In the field of digital libraries, CADAL has accumulated a large number of digital books, and provides the corresponding book reading service. Obviously, each book contains a large number of pieces of knowledge. It will generate greater value to break down, associate and recombine these resources by specific target.So, in the base of CADAL’s massive digital library resources, we complete the book knowledge extraction and decomposition by building a Concept-Focus Knowledge Network and complete the association and organization by generating a concept entry. Our contribution are as follows:(1) The construction of Book based Concept-Attribute Knowledge Network. First, we present the prototype of the knowledge network. Then, we propose a word embedding based matching model to extract chunk in book catalog and classify trunk to concept and attribute by using SVM. Finally, we get the knowledge network.(2) We propose a Learning to Rank based algorithm to generate the catalog of a concept entry. The proposed algorithm sorts attribute candidates, which retrieved from Concept-Attribute Knowledge Network, to reflect the knowledge timeline.(3) We design and implement an automatic generation system for concept entries based on a Concept-Attribute Knowledge Network.