Алгоритм обработки информации для прогнозирования развития технологий на примере теле- и радиовещания

Abstract

Today, many Russian companies working in the field of software development for television and radio broadcasting can extract the necessary data on trends in their field of activity by analyzing information received from foreign competitors, for example, studying specialized popular science literature posted on the official websites of foreign competitors, when analyzing information obtained from exhibitions and presentations. However, the opportunity to comprehensively analyze the novelties of foreign developers patented in the world patent databases will give Russian companies competitive advantages in the formation of strategic development plans. It is proposed to develop a patent landscape based on the data of the global patent array (in English) in the field of television and radio broadcasting, analyze patent trends and predict the development of technologies in this area. The novelty of the developed method for predicting the development of technologies lies in the cooperation of USPTO (USA) patent clustering methods based on the international patent classification and forecasting technological trends using the generated time series. The developed method is implemented in Python; in parallelizing the process of parsing patents, the Multiprocessing library was used, Statsmodels was used to analyze data and carry out the forecasting process, and Matplotlib was used to visualize data. The MySQL database management system was chosen to organize information storage. There were processed 28,591 patents, the process of forecasting the development of television and radio technologies was tested based on the constructed time series in the interval of 2015–2020.