KeyWise AI is a developer of data-driven AI software that turns smartphones into Fitness Trackers for Brain Health.
Our technology is built on a deep foundation of extensive pilot test findings and peer-reviewed publications 1-6, as a result of our ongoing BiAffect open-science project: https://www.biaffect.com. The BiAffect project demonstrated that the way we interact with our smartphone, centered around keyboard dynamics, provides insights into our brain health. Such keyboard interaction dynamics can passively and unobtrusively collect data from our real-world behaviors as they occur in everyday life. Such a continuous pipeline of real-world data can then be used to derive metrics on how our brains are functioning.
KeyWise AI was built on the extensive experience working with keyboard metadata as part of the BiAffect study.
KeyWise AI was founded as a university spin-off building on the generalizable knowledge that our team has learned from running the highly successful BiAffect crowd-sourced research study. Leveraging Apple’s ResearchKit framework to enroll and study eligible U.S. adults directly via their personal iPhones, the BiAffect iOS research app has enrolled more than 2,000 citizen scientists, collected 23,000+ hours of smartphone keyboard dynamics data that resulted in 32M+ keystrokes.
The rich dataset obtained from BiAffect allowed the KeyWise team to develop the 4 metrics included in the KeyWise app: Processing Speed, Attention, Impulse Control, and Mood Stability. These metrics use information from your personal typing data (e.g., speed of typing, number of daily keypresses, use of backspace, use of autocorrect, etc.) and state-of-the-science on-device machine learning to derive real-time data-driven metrics on how your brain is functioning. We take your privacy seriously, and the words you type are not captured.
You’ll notice that you need to contribute your typing data takes from 1 to 5 days for KeyWise to have enough keystroke data to calculate your brain health metrics. During this time, you are also adapting to using a new keyboard on your phone, and your use of the new keyboard will improve over time. Our algorithms are continuously learning your typing patterns, so the more keystroke data you provide the more accurate your metrics will be.
Just as monitoring your step count is important for seeing how active you are and improving your physical health, monitoring these KeyWise AI metrics can be helpful for brain health. Paying attention to these metrics may encourage you to engage in activities that support brain health like exercise, good sleep, and stress management. With KeyWise AI, you can also monitor how well these behavioral health activities improve brain health metrics such as mood stability and attention.
- Zulueta J, Piscitello A, Rasic M, et al. Predicting Mood Disturbance Severity with Mobile Phone Keystroke Metadata: A BiAffect Digital Phenotyping Study. Journal of medical Internet research. 2018;20(7):e241.
- Stange J, Zulueta J, Langenecker S, et al. Let your fingers do the talking: Passive typing instability predicts future mood outcomes. Bipolar disorders. 2018;20(3):285-288.
- Cao B, Zheng L, Zhang C, et al. DeepMood: Modeling mobile phone typing dynamics for mood detection. Proceedings of the 23rd ACM SIGKDD International - Conference on Knowledge Discovery and Data Mining; 2017; Halifax, NS, Canada.
- Vesel C, Rashidisabet H, Zulueta J, et al. Effects of mood and aging on keystroke dynamics metadata and their diurnal patterns in a large open-science sample: A BiAffect iOS study. J Am Med Inform Assoc. 2020;27(7):1007-1018.
- Sun L, Wang Y, Cao B, Yu PS, Srisa-an W, Leow AD. Sequential Keystroke Behavioral Biometrics for Mobile User Identification via Multi-view Deep Learning. In: Machine Learning and Knowledge Discovery in Databases. Springer; 2017:228-240.
- Huang H, Cao B, Yu PS, Wang CD, Leow AD. dpMood: Exploiting local and periodic typing dynamics for personalized mood prediction. IEEE International Conference on Data Mining (ICDM); 2018; Singapore.