R&D proposal screening system based on text-mining approach
Changwoo Choi, Yongtae Park · International Journal of Technology Intelligence and Planning · 2006
With the enormous accumulation of R&D projects, it becomes very difficult to avoid their duplication; thus, the screening process of R&D proposals must rest on intelligent systems, rather than on intuitive criteria. This is particularly obvious for large-scale, multi-product firms that deal with multi-technology R&D projects. However, previous research on R&D knowledge management has focused on the ex post facto project evaluation/selection stage, instead of the preliminary screening stage. In this research, we propose an intelligent and systematic R&D proposal screening system that may be employed to identify and filter out duplications of previous or ongoing projects. The screening algorithm is based on the coherence of contents and substance, not on the coincidence of titles, between a new R&D proposal and the pool of completed and existing R&D projects. The text-mining methodology is employed to this end. An illustrative set of real data is applied to elucidate the detailed procedure of the system.