• named entity recognition
• NER
• patents
• diabetes device
• text mining
• information retrieval

Patents contain important information that can be essential for different stakeholders, such as Policy makers, researchers and private companies. If analyzed carefully, they can show technological details and relationships, reveal business trends, inspire new industrial solutions or help in making investment policies. This thesis aims to use these documents as a tool to map concepts related to the medical field. Especially in the case of chronic diseases such as diabetes, the development of technologies that can improve their management is of particular importance. Many steps forward have been made, incorporating concepts such as Industry 4.0 and IoT within medical inventions. Through the study of patents for diabetic devices, it was possible to analyze the technological developments made, which are increasingly directed towards the use of smart technologies. Text mining tools were then used to extract three types of entities: users mentioned in the inventions, diabetes-related diseases and technologies related to Industry 4.0. The analysis of these three entities made it possible to understand which devices are being researched, which diseases are being focused on and which types of users will be targeted.