A HYBRID APPROACH TOWARDS INFORMATION EXPANSION BASED ON SHALLOW AND DEEP METADATA
Tudor Groza, Siegfried Handschuh · 2009
The exponential growth of the World Wide Web in the last decade, brought an explosion in the information space, with important consequences also in the area of scientific research. Lately, finding relevant work in a particular field and exploring links between relevant publications, became a cumbersome task. In this paper we propose a hybrid approach to automatic extraction of semantic metadata from scientific publications that can help to alleviate, at least partially, the above mentioned problem. We integrated the extraction mechanisms in an application targeted to early stage researchers. The application harmoniously combines the metadata extraction with information expansion and visualization for the seamless exploration of the space surrounding scientific publications.