| Abstract |
A transformation process is any activity or group of activities that takes some inputs such as materials and energy, and transforms it into one or several outputs such as products and services. There has been a growing interest in discovering knowledge from transformation processes, especially from the more complex ones, in order to reduce their cost, improve their outputs' quality, and reduce the environmental impact caused by these processes. Depending on the objectives, the final knowledge can be a new result, a discovery of unknown information, a prediction, or a confirmation about consolidated facts and theories. In this talk, I present our approach of discovering knowledge, and in particular detecting patterns from scientific datasets, based on a formal knowledge representation of the transformation processes, and the detection of contextual identity relationships between the participants of the processes. |