Mining Technical Knowledge

The information field companies are living in has changed dramatically over the last years bringing a new challenge for management engineers. This discipline comes with engineering methodologies applied to inherent systems but, nowadays, activities with greater added value for companies are hardly standardized and non-repetitive. The enormous amount of information, which is changing the environment of companies, has a determinant impact on Research and Development, Design, Marketing and Human Resources Management: all functions with high strategic content, and so knowledge. Since documents written in natural language contains knowledge by design, management engineers has nowadays the great opportunity to exploit the technical knowledge hidden in this unstructured sources to generate value. The aim of this thesis is to design methods and processes for the analysis of technical documents in order to extract valuable knowledge for companies. The methods are ensembles of Natural Language Processing and Managements Engineering techniques. The methods has the goal of providing correct knowledge exchange between humans and machines, leading to incorporate knowledge of the experts inside machine-learning systems and experts’ ability to use in their process of decision making inductively generated knowledge of machines.