Discovering technology and science innovation opportunity based on sentence generation algorithm

Abstract

Science and technology are crucial elements in discovering innovation opportunities. They have their own practical and theoretical unique meaning in innovation factors. Scientific or techno-logical information can be collected by patents or papers, and various approaches for innovation opportunities discovery are being proposed using text mining. Since a previous methodology using patents and papers discovered opportunities in science or technology itself, they cannot discover opportunities reflecting the science and technology relationship. In addition, since dis-covered innovation opportunities are formed within the keyword or phrase level, they cannot provide innovation direction or purpose. Therefore, this study suggests a new approach to dis-covering science and technology innovation opportunities that reflects the science-technology relationship and their concrete directions/purposes using the sentence generation algorithm. An algorithm-generated sentence can contain contextual flow and connection between keywords. In contrast, the generated sentences from the sentence model can reflect science and technology from mass data in a readable sentence. Key innovation factors from science and technology are extracted from generated sentences and then innovation opportunities with specific directions and purposes are suggested.