Less Data, More Research

The idea of my presentation is to demonstrate a QoS aware information retrieval framework that will complement Apple’s ResearchKit. This will aid the medical research community to increase the degree of participation among iPhone user groups and capture more significant and relevant data. (https://developer.apple.com/researchkit/researchkit-technical-overview.pdf). The success of the proposed ResearchKit widely depends on the users’ participation and willingness to share large amount of information collected via mobile phone sensors. My proposed research kit helper will provide probabilistic data aggregation and query processing filters to minimise network traffic while still maintaining the quality of survey results. A number of quality adjustable query processing filters will be prebuilt as part of my ResearchKit helper and will be configurable via backend. The core features and strengths of the proposed ResearchKit Helper framework: - It will provide basic data processing filters. Researchers are not interested in bulk data (motion sensor or activity data from phone), rather statistically significant representative data will be produced. - A number of generic probabilistic models will be pre-built (e.g. Merge distributions) - This will enable the research community to engage larger audience. - Degree of participation will be higher as this will minimise the data traffic for research survey - Configurable data processing filters that can be adjusted from a remote backend. - Proposed framework will also facilitate anonymization of data for protecting user’s privacy

This talk was recorded at AltConf 2015. Watch all the videos!