Combining Machine Learning with a Pharmaceutical Technology Roadmap to Analyze Technological Innovation Opportunities.

Technology roadmaps (TRM) with flexible architectural structures and development processes can effectively mitigate innovation dilemmas in pharmaceutical technology innovation opportunity analysis, such as long innovation cycles and inefficient translation. However, pharmaceutical technology innovation has become complicated and increasingly diverse. The current TRMs do not adapt well to this new reality. In response, this study proposes a systematic and specific framework to develop a pharmaceutical technology roadmap. Compared with standardized TRM, this TRM proposes three improvements. The first extension is the designing of the layers. The second extension is the selecting of the data source. The last extension is the processing of data sources. To validate the proposed framework, a case study in the field of hyperuricemia drugs was conducted. Our analysis focuses on the construction of a pharmaceutical TRM framework and a review of technology topics from multiple data sources that can supply quantitative information to explore the current trends. The analyzed results can assist the research and development professionals to predict technological opportunities and support experts to make more reasonable decision-making for a particular pharmaceutical domain.