• advantage of an invention
• drawback of an invention
• named entity recognition
• niche market
• patent analysis
• text mining
• User of an invention

In this thesis is discussed a novel approach to extract unstructured information from patents. Using specific Classifiers is possible to extract three relevant entities of the invention, described in a patent: Users, Drawbacks and Advantages.
The information extracted is addressed to designers and marketers that are involved in the process of collection of data during the identification of new Users and during the New Product Development process.
The first part of this thesis describes the Classifiers and the two approaches used to validate the extracted entities. In the second part, the analysis is focused on Users and on the potentialities of this novel approach to discover generic Users or niches in several technological fields. The analysis concerns patents on oral care devices.