Patent-based inventor profiles as a basis for human resource decisions in research and development

Profiles of inventors' technological competence are a valuable source of information for decision-making in research and development (R&D) management, e.g. concerning inventor assessment, human resource development and R&D team-building. In the following exposition, a new method of inventor profiling will be put forward, which is based in particular on semantic patent analysis and multidimensional scaling. First, in the course of semantic patent analysis, specialized software, equipped with a natural language processor, reads the patent text transferring the contents into a subject-action-object-format (SAO). The extracted SAO structures are then used to create similarity matrices for patents or patent sets, respectively, according to a specific similarity value. Subsequently, an inventor competence map can be produced by means of multidimensional scaling. The benefits of this method for R&D-related issues in human resource management will be illustrated by the example of a German mechanical engineering company. Two distinct types of profiles were generated and tested: (i) the profile of a single key inventor and (ii) a profile of key inventor sets. The single key inventor profile gives information on the range of competence, i.e. the homogeneity or heterogeneity of a certain inventor's competences, providing far more detailed insights than resorting to bibliographic data like international patent classification (IPC) classes or citations, whereas the latter kind of profile establishes the position of a certain key inventor in relation to others, helping to highlight specific groups of inventors and their domains. These results are clearly apt to support human resource management.