Major developed countries, including Korea, commit tremendous efforts to increase the utilization of research outcomes from the government-led R&D programs. Typical research outcomes include papers, patents, and R&D reports. However, most domestic and foreign services just provide keyword search services only for metadata. In addition, there is a problem where it takes too much time to read the retrieved full-text data and check whether the result is a desired result. Therefore, there is a necessity of service advancement.

In this paper, we developed an XML DTD which converted the R&D report in PDF format into XML format, and constructed a DB by indexing each element using a search engine. In addition, non-text (table, figure) is extracted automatically in the conversion process when DB is constructed. In this way, if the need of keyword search rises, one can search in three ways: metadata search result, full-text page-based search result, and non-text (table, picture) search result. We proposed a service that can select only the desired results among the three search results and store them in a single PDF file.

R&D report, XML, retrieval system, bundle service.