Title: Principles, Algorithms, and Interfaces of Touchscreen Text Entry Systems

Abstract:
Text entry is one of the most basic, common and important tasks on Post-PC computing devices such as smartphones and tablets. However, entering text on a touchscreen keyboard is inherently error-prone and challenging, because of the imprecision of finger touch and small key sizes. In this talk, I will introduce how I and my collaborators have been working to address this challenge. First of all, we have derived mathematical models and principles for designing the modern smart touchscreen keyboard, which is able to correct and predict the words a user intends to enter based on the noisy and incomplete spatial signals from touch input, and the knowledge from language models. Applying rigorous mathematical optimization methods, we have designed a keyboard that is able to correct 70% of user’s erroneous input and save 37% of keystrokes. Second, to allow users to gesture words while holding a device in two hands, we have designed and implemented the bimanual gesture keyboard, extending the gesture typing paradigm from one finger to multiple fingers. Third, we explored the optimality space for keyboard layout design. Our effort led to new layouts more efficient and accurate than Qwerty. We also investigated how to improve the learnability of the optimized layouts. From these innovations, we hope to provide users efficient, intelligent, and natural text entry experiences on various touchscreen devices.

Brief Biography:
Xiaojun Bi joined the Department of Computer Science at Stony Brook University as an Assistant Professor in January 2017. Previously, he was a Research Scientist at Google in Mountain View, California. His research interests center on Human-Computer Interaction (HCI) in Post-PC computing, with major focuses on interactive system building, algorithm design and human behavior modeling. Some of his research including virtual keyboard correction and completion algorithm and bimanual gesture typing algorithm has been integrated into the Google keyboard, making real world impacts and benefiting millions of users. He earned his Ph.D. from the Department of Computer Science at the University of Toronto, and master’s and bachelor’s degrees from Tsinghua University. Xiaojun Bi has over 20 publications in the top tier HCI conferences and journals including CHI, UIST and Human Computer Interaction, and is an inventor/co-inventor of over 20 US patents. He constantly serves as a program committee member at premier HCI conferences such as CHI and UIST . More information can be found on his homepage: http://www.xiaojunbi.com/