Detecting Technological Opportunities Using Technical Keywords from Patents (ICMIT 2015)

Technological opportunities are an important factor for innovative outputs in most companies. To detect technological opportunities, patent analysis has been regarded as a basis for technology assessment in this paper. With too much expert dependence in previous research, this paper suggests a reliable way to detect technological opportunities using technical keywords from patents. Using a keyword-based model, technical keywords are decided by term frequency-inverse document frequency (TF-IDF) heuristics. After the relatedness of technical keywords is established by distributional similarity, patent documents are clustered using a hierarchical clustering algorithm through semantic processing. Technological opportunities can be identified by estimating clusters. To prove this, a case of electrochemical glucose biosensors is given.