
Brain storming

Prototying

Presentation and Discussion

we propose a mobile crowd sourcing platform, MoTurk, for transactions of mobile sourced personal data between the general public and researchers through a Mechanical Turk mechanism.
we mainly reconcile the dilemma of flexible acquisition of mobile sensing data and privacy protection by summarizing features that support a four layer hierarchical model and implementing a functional programming model for data processing.
Mobile Phones are just every where. It unobtrusively collects various of our personal data.These kinds of personal data are of great research needs in various research areas. For, While Sleeping hours are strongly related to clinic health and academic performance, a better understanding of travel pattern can largely contribute to urban planning.
However, current implementation of mobile sensing data for research use encounters two major difficulties:
Data acquisition difficulty:Researchers typically need to develop a data collector app for collecting mobile sourced data(e.g.,coordinate, audio data). Moreover, as raw sensing data can hardly be used as behavior indications, further data transformation and processing methods are required.
Privacy concern: Current 'Informed Consent' policy provides participants with less power of privacy control over how and to what extent their personal data are used. There also lack references for tangible value of personal data, making it hard for the public to decide whether or not to share their privacy.
MoTurk attempts to reconcile the dilemma between data acquisition and privacy protection via:
(1)Online customization; (2)Feature extractor; (3)Privacy enforcement
Literature Review
Data Strictire Construction
Paper writting
MoTurk was done in HCII,CMU. Many thanks to the guidance of Fanlin Chen and Prof. Jason Hong.