Technological Opportunities Discovery for Safety Through Topic Modeling and Opinion Mining in the Fourth Industrial Revolution: The Case of Artificial Intelligence.

In this article, the fourth industrial revolution (4th IR) has features that are distinct from the previous three industrial revolutions in terms of a paradigm shift. Although 4th IR relevant technologies may have positive impacts in many fields, they may also create harmful impacts, just as every previous industrial revolution has. Most studies dealing with the 4th IR have focused on its positive impacts; the present research concentrates on its negative impacts and suggests technological opportunities to hedge against expected risks in the future. To this end, future societal topics are first identified through latent Dirichlet allocation based topic modeling using textual information of futuristic data, which is retrieved from the website, including information on future-related insights. Second, the risk topics related to the future are selected with the help of a weighting score using predefined risk keywords and sentiment analysis. Finally, safe technological opportunities are identified based on textual information derived from patent data related to risk topics with the perspectives on risk analysis that have prior (prediction, prevention) and posterior perspectives (response, recovery). To illustrate technology relevant to the 4th IR, artificial intelligence (AI) is selected. As a result, 36 topics related to future society and six future risk-related topics are suggested that reflects security-relevant properties of AI and three types of threats. Opportunities to ensure or safety through technology are also proposed in four perspectives. Technological opportunities related to safety are explored in two perspectives. One is to expand existing threats, and the other is to suggest new threats. This research can be utilized to plan future safety technology by taking a data-driven approach under the huge paradigm shift of the 4th IR and proposing potential technological solutions that are not future societal scenarios and policy solutions suggested in the previous studies. [ABSTRACT FROM AUTHOR]