PRIDE (Personal Risk Detection) dataset

We define personal risk detection as the timely identification of when someone is in the midst of a dangerous situation, for example, a health crisis or a car accident, events that may jeopardize a person’s physical integrity. We work under the hypothesis that a risk-prone situation produces sudden and significant deviations in standard physiological and behavioural user patterns. These changes can be captured by a group of sensors, such as the accelerometer, gyroscope, and heart rate.

The PRIDE (Personal Risk Detection) dataset is built with the help of 18 test subjects and a period of data collection of one week each, 24 h per day; the normal conditions dataset (NCDS) is built in this manner. Next, to build the anomaly conditions dataset (ACDS), the same 18 test subjects collaborate in another process to gather data under specific abnormal or stressful conditions. The scenarios include activities such as running several meters, going up and down the stairs of a several-floor building as fast as possible, simulating a fight, and so on.

 — Dataset usage terms : The use of the data is only for scientific purposes, please provide the corresponding credit to the owners when publishing any work based on this data

Please cite:

Ari Yair Barrera-Animas, Luis A. Trejo, Miguel Angel Medina-Pérez, Raúl Monroy, J. Benito Camiña, and Fernando Godínez. 2017. Online personal risk detection based on behavioural and physiological patterns. Information Sciences 384, (2017), 281–297. doi: 10.1016/j.ins.2016.08.006

For more information, please read:

• Jorge Rodríguez, Ari Y. Barrera-Animas, Luis A. Trejo, Miguel Angel Medina-Pérez, and Raúl Monroy. 2016. Ensemble of One-Class Classifiers for Personal Risk Detection Based on Wearable Sensor Data. Sensors 16, 10 (2016). doi: 10.3390/s16101619
• Luis A. Trejo and Ari Yair Barrera-Animas. 2018. Towards an Efficient One-Class Classifier for Mobile Devices and Wearable Sensors on the Context of Personal Risk Detection. Sensors 18, 9 (2018). doi: 10.3390/s18092857
• Miryam Elizabeth Villa-Pérez and Luis A. Trejo. 2020. m-OCKRA: An Efficient One-Class Classifier for Personal Risk Detection, Based on Weighted Selection of Attributes. IEEE Access 8, (2020), 41749–41763. doi: 10.1109/ACCESS.2020.2976947