Keywords: Shiny, JavaScript, Data Collection, User Experience
Webpages:
https://github.com/nstrayer/shinysense,
https://nickstrayer.shinyapps.io/shinysense_earr_demo/,
https://nickstrayer.shinyapps.io/shinysense_swipr_demo/ shinysense is a package containing
shiny modules all geared towards helping users make mobile-first apps for collecting data, or helping
Shiny “sense” the outside world. Currently the package contains modules for gathering data on swiping (shinyswipr), audio (shinyearr), and from accelerometers (shinymovr). The goal of these functions is to take
Shiny from a tool for demonstrating finished models or workflows into being a tool for data collection, enabling its use for training/testing models or building richer user experiences.
Several demo apps are contained in the package, including training and testing a speech recognition system using shinyearr and detecting spell casts performed by swinging your phone like a wand with shinymovr. In addition, the package is already being used in real-world products. Notable examples being the app
papr which allows users to rapidly read and react to abstracts by swiping cards containing their content using shinyswipr, validating algorithm output in
GenomeBot Tweet Generator, and
contributr: an app that allows users to review github issues on various R packages.
A major goal of the construction of
shinysense was mobile-friendly behavior. The massive proliferation of smartphones laden with sensors is a potential goldmine of data and use cases for statisticians and data scientists. This package attempts to help users harness this new flood of opportunities. A side effect of mobile oriented design is increased usability in non-static environments (Dou and Sundar 2016, Wigdor, Fletcher, and Morrison (2009)). For instance: an app running on a smartphone allowing physicians to input parameters into clinical models and instantly see the results (shinyswipr), to the generation and testing of fitness tracking algorithms by carrying a phone in a pocket (shinymovr).
Much in keeping with the primary goal of
Shiny, by bringing powerful software tools such as inputs typically reserved for
JavaScript/ native apps to a tool used by scientists such as
R we hope to lower the costs (monetary and otherwise) of bringing innovative applications to fruition.
References Dou, Xue, and S Shyam Sundar. 2016. “Power of the Swipe: Why Mobile Websites Should Add Horizontal Swiping to Tapping, Clicking, and Scrolling Interaction Techniques.”
International Journal of Human-Computer Interaction 32 (4). Taylor & Francis: 352–62.
Wigdor, Daniel, Joe Fletcher, and Gerald Morrison. 2009. “Designing User Interfaces for Multi-Touch and Gesture Devices.” In
CHI’09 Extended Abstracts on Human Factors in Computing Systems, 2755–8. ACM.
