RL-KLM: automating keystroke-level modeling with reinforcement learning

The Keystroke-Level Model (KLM) is a popular model for predicting users' task completion times with graphical user interfaces. KLM predicts task completion times as a linear function of elementary operators. However, the policy, or the assumed sequence of the operators that the user executes, needs to be prespeciffed by the analyst. This paper investigates Reinforcement Learning (RL) as an algorithmic method to obtain the policy automatically. We define the KLM as an Markov Decision Process, and show that when solved with RL methods, this approach yields user-like policies in simple but realistic interaction tasks. RL-KLM offers a quick way to obtain a global upper bound for user performance. It opens up new possibilities to use KLM in computational interaction. However, scalability and validity remain open issues.

[1]  Alex Graves,et al.  Playing Atari with Deep Reinforcement Learning , 2013, ArXiv.

[2]  John E. Laird,et al.  Soar-RL: integrating reinforcement learning with Soar , 2005, Cognitive Systems Research.

[3]  I. Scott MacKenzie,et al.  Fitts' Law as a Research and Design Tool in Human-Computer Interaction , 1992, Hum. Comput. Interact..

[4]  Antti Oulasvirta,et al.  Informing the Design of Novel Input Methods with Muscle Coactivation Clustering , 2015, ACM Trans. Comput. Hum. Interact..

[5]  Stuart J. Russell Learning agents for uncertain environments (extended abstract) , 1998, COLT' 98.

[6]  Harold Thimbleby,et al.  Usability analysis with Markov models , 2001, TCHI.

[7]  Martin L. Puterman,et al.  Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[8]  Michael D. Byrne,et al.  ACT-R/PM and menu selection: applying a cognitive architecture to HCI , 2001, Int. J. Hum. Comput. Stud..

[9]  Jan Peters,et al.  Reinforcement learning in robotics: A survey , 2013, Int. J. Robotics Res..

[10]  José A. Pino,et al.  Predicting user performance time for hand gesture interfaces , 2017 .

[11]  Marti A. Hearst,et al.  The state of the art in automating usability evaluation of user interfaces , 2001, CSUR.

[12]  R. Hyman Stimulus information as a determinant of reaction time. , 1953, Journal of experimental psychology.

[13]  Jieun Kim,et al.  Fingerstroke time estimates for touchscreen-based mobile gaming interaction. , 2015, Human movement science.

[14]  Allen Newell,et al.  The keystroke-level model for user performance time with interactive systems , 1980, CACM.

[15]  David E. Kieras,et al.  The GOMS family of user interface analysis techniques: comparison and contrast , 1996, TCHI.

[16]  Antti Oulasvirta,et al.  Modelling Learning of New Keyboard Layouts , 2017, CHI.

[17]  Antti Oulasvirta,et al.  The Emergence of Interactive Behavior: A Model of Rational Menu Search , 2015, CHI.

[18]  David E. Kieras,et al.  An Overview of the EPIC Architecture for Cognition and Performance With Application to Human-Computer Interaction , 1997, Hum. Comput. Interact..

[19]  Marisa E. Campbell,et al.  Conference preview: HCI 2000: usability or else! , 2000, INTR.

[20]  Carl Gutwin,et al.  Storyboard-Based Empirical Modeling of Touch Interface Performance , 2018, CHI.

[21]  John Millar Carroll HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science , 2003 .

[22]  Antti Oulasvirta,et al.  Performance and Ergonomics of Touch Surfaces: A Comparative Study using Biomechanical Simulation , 2015, CHI.

[23]  Mark D. Dunlop,et al.  Enhancing KLM (keystroke-level model) to fit touch screen mobile devices , 2014, MobileHCI '14.

[24]  Albrecht Schmidt,et al.  Keystroke-level model for advanced mobile phone interaction , 2007, CHI.

[25]  I.,et al.  Fitts' Law as a Research and Design Tool in Human-Computer Interaction , 1992, Hum. Comput. Interact..

[26]  Michael D. Byrne,et al.  A tool for creating predictive performance models from user interface demonstrations , 1999, UIST '99.

[27]  Roberto Pieraccini,et al.  A stochastic model of human-machine interaction for learning dialog strategies , 2000, IEEE Trans. Speech Audio Process..

[28]  Kwang-Ting Cheng,et al.  Automatic Functional Test Generation Using The Extended Finite State Machine Model , 1993, 30th ACM/IEEE Design Automation Conference.

[29]  Steven C. Seow Information Theoretic Models of HCI: A Comparison of the Hick-Hyman Law and Fitts' Law , 2005, Hum. Comput. Interact..