Unsupervised Metadata Extraction in Scientific Digital Libraries Using A-Priori Domain-Specific Knowledge.
Alexander Ivanyukovich, Maurizio Marchese · 2006
Abstract — Information extraction from unstructured sources is a crucial step in the semantic annotation of content. The challenge is in supporting an high quality automatic approach (or at least semi-automatic) in order to sustain the scalability of the semantic-enabled services of the future. Unsupervised information extraction encompasses a number of underlying research problems, such as natural language processing, heterogeneous sources integration, knowledge representation, and others that are under past and current investigation. In this paper we concentrate on the problem of unsupervised metadata extraction in the Digital Libraries domain. We propose and present a novel approach focusing on the improvement in the metadata extraction quality without involving external information sources (oracles, manually prepared databases, etc), but relying on the information present in the document itself and in its corresponding context. More specifically, we focus on quality improvements of metadata extraction from scientific papers (mainly in computer science domain) collected from various sources over the Internet. Finally, we compare the results of our approach with the state of the art in the domain and discuss future work. I.