Data-Oriented Software Development: The Industrial Landscape through Patent Analysis

Abstract

Τhe large amounts of information produced daily by organizations and enterprises have led to the development of specialized software that can process high volumes of data. Given that the technologies and methodologies used to develop software are constantly changing, offering significant market opportunities, organizations turn to patenting their inventions to secure their ownership as well as their commercial exploitation. In this study, we investigate the landscape of data-oriented software development via the collection and analysis of information extracted from patents. To this regard, we made use of advanced statistical and machine learning approaches, namely Latent Dirichlet Allocation and Brokerage Analysis for the identification of technological trends and thematic axes related to software development patent activity dedicated to data processing and data management processes. Our findings reveal that high-profile countries and organizations are engaging in patent granting, while the main thematic circles found in the retrieved patent data revolve around data updates, integration, version control and software deployment. The results indicate that patent grants in this technological domain are expected to continue their increasing trend in the following years, given that technologies evolve and the need for efficient data processing becomes even more present.