Capturing weak signals of technology-driven future

Abstract

This thesis aims to propose a novel procedure which is capable of detecting weak signals of specific technology and interpreting them. As technology innovation nowadays brings discontinuous transformations to human life, the importance of forecasting discontinuities has been increased. But the usefulness of existing forecasting method, which highly depends on cumulative knowledge of experts, has been decreased. As an alternative to this, weak signal analysis for exploring future opportunities has been getting attention. However, there are limitations of existing weak signal analysis which is qualitatively conducted by a small group of experts

not only labor-intensive and too costly, but also hard to find various weak signals and to interpret them. In response, this paper suggests a new approach that detects weak signal quantitatively from web data and helps generate scenarios with weak signals. In the process, IdeaGraph algorithm, which is text mining algorithm of summarizing document into keyword network based on co-occurrence of keywords, is revised and utilized. Rare and significant keywords can be extracted using IdeaGraph, and they can be identified as weak signals. Furthermore, information about which keywords are highly related to identified weak signals can be obtained and it is helpful to scenario generation. Proposed weak signal analysis procedure is composed of four steps

(1) Constructing database by collecting and preprocessing web documents from selected websites saying about technology being concerned, (2) developing keyword network and finding impactful keywords using revised IdeaGraph algorithm, (3) identifying weak signals among impactful keywords by filtering high-frequency terms, and (4) generating scenarios with weak signals utilizing information from developed keyword network. As an illustrative example, proposed procedure is applied to virtual reality technology. Proposed approach enables not only detecting weak signal quickly and effectively based on quantitative criteria, but also generating scenarios with weak signals systematically. With the help of proposed procedure, it is expected that various organizations would be able to find opportunities caused by technology innovation in advance and to establish an effective strategy based on the discovered opportunities.