?
Использование семантического анализа для автоматизированного выявления хайповых технологий
The research focuses on inflated public expectations of new technologies, or hypes. The paper presents the results of the development and testing an automated methodology for the automated identification of hypes among technological topics based on their textual trace in the digital technology field. The amount of new technological developments in the world is constantly growing, however their real potential for practical application varies greatly. Therefore, it is important to understand reliable factors to distinguish trends from hypes. Typically, industry and technology experts suggest that the possible signs of hypes include the following ones: absence of a stable business model, an unformed or obviously limited consumer market, and a large number of more effective alternatives. Identifying hypes in the technology agenda remains a difficult analytical task. This is due to the terminological inconsistency, the expert nature of the task, insufficiently developed methodological approaches, and the lack of specific technical tools. The method described in this paper involves the extraction of terms referring to technologies using natural language processing and computational linguistics techniques. These terms are extracted from several dozens of millions of different types of text documents, such as scientific publications, patents, and market analytics. The method also includes calculating an objective measure of each technology's “hype” and constructing a visual map that illustrates the technology landscape that allows separating sustainable trends from potential hypes. Decision makers can use such hype maps in conjunction with other analytical results to identify priority development areas, analyze current and forecast future trends, and in risk management.