Expanding Sensor Networks to Automate Knowledge Acquisition

摘要: The availability of accurate, low-cost sensors to scientists has resulted in widespread deployment a variety sporting and health environments. sensor data output is often raw, proprietary or unstructured format. As result, it difficult query multiple for complex properties actions. In our research, we deploy heterogeneous network detect the various biological physiological athletes during training activities. goal exercise physiologists quickly identify key intervals such as moments stress fatigue. This not currently possible because low level lack language support. Thus, motivation expand with contextual layer that enriches raw data, so can be exploited by high language. To achieve this, domain expert specifies events tradiational event-condition-action format deliver required enrichment.