NERMAP: Towards Automating Technology Roadmapping

In the process of creating a Technological Roadmap, a large amount of data has to be analyzed and future events may be spread across several document databases, making it unlikely to researchers retrieve all future events described in these publications in a timely fashion. Technology Roadmapping is a time-consuming and error-prone method, which requires a group of experts to conduct the study. This work proposes a system capable of semi-automating the Technology Roadmapping Process, through the Named Entity Recognition technique supported by machine learning, retrieving up to 83% of future events in documents when compared to humans. Furthermore, the system was able to identify some future events that our validation team missed during our tests. Researchers using the system can perform the analysis in considerably less time, reducing associated costs. Therefore, the proposed system enables a small group of researchers to analyze a large mass of documents in a short time, streamlining their ability to identify, process, and analyze information in the form of roadmaps.