SOCIOECONOMIC INDICATORS DETERMINING INNOVATION: ASSOCIATION BASED ON DATA MINING

Abstract

The sustained growth of an economy is linked to innovations in various sectors. An analysis based on economic indicators can guide the aspects of the economy to be prioritized for the creation of an environment conducive to innovation in a country. Thus, this study analyzed the most relevant socioeconomic aspects in constructing an innovative environment with the contribution of data mining through the J48C algorithm, demonstrating how data mining can contribute to the development of models for governmental decision-making and the construction of public policies. A total of 119 socioeconomic indicators were used, and 32 countries available in the innovation ranking of the Global Innovation Index (GII) 2020 report were selected. For data processing, the decision tree technique was adopted using the WEKA software and based on the C4.5 algorithm. The countries include the most and least innovative by region, which were processed via data mining through the supervised J48C algorithm. To highlight the socioeconomic indices most related to the more or less innovative outcome, a separation of the indicators by Pillars based on the Global Innovation Index (GII) 2020 report was conducted. The results highlight demographic indicators, labor availability, production of goods of both high and medium technology, and governmental efficiency in association with more innovative economies. The relevance of women's participation in business and the relationship between female entrepreneurship and innovation are also highlighted. Most related research with pillar and indicator contributed to the findings and validated data mining as a method to investigate issues related to the development of economies through innovation.