12.3 - A Process Model for the Discovery of Knowledge in Sensor-Based Indoor Climate Data
- Event
- ettc2018 - European Test and Telemetry Conference
2018-06-26 - 2018-06-28
Nürnberg, Germany - Chapter
- 12. Data Management Applications
- Author(s)
- D. Gruedl, T. Wieland - Fraunhofer Anwendungszentrum Drahtlose Sensorik, Coburg (Germany)
- Pages
- 293 - 299
- DOI
- 10.5162/ettc2018/12.3
- ISBN
- 978-3-9816876-7-5
- Price
- free
Abstract
Sensors in office spaces collect data about the indoor climate in order to monitor and control HVAC-systems to ensure a constant air quality. The quality of indoor climate is dependent on various aspects, such as carbon dioxide, temperature, and humidity. Approaches on searching patterns in this sensor data by using data mining techniques rarely explicitly apply established process models like KDD (knowledge discovery in databases) or CRISP-DM (cross industry standard process for data mining). These process models describe data mining as one of multiple steps in the whole process of extracting patterns from data. Other steps include the preprocessing of the data as well as the evaluation of the extracted patterns after the data mining. This paper analyzes these aforementioned approaches and compares them to the established process models before deriving a process model that can be widely applied to data mining projects searching for patterns in sensor-based indoor climate data. The derived process model puts more emphasis on understanding the data and its context as preliminary steps to the extraction of patterns via data mining. In addition to facilitating new research on indoor climate data, the derived process model also allows to understand research in this field of study more easily.