In collaboration with Paulson School of Engineering and Applied Sciences, Harvard University
Reuth Mirsky, Ya’akov Gal, Stuart Shieber
ACM Transactions on Intelligent Systems and Technology 8 (3), 2017, 45
In exploratory domains, agents’ behaviors include switching between activities, extraneous actions, and mistakes.Such settings are prevalent in real world applications such as interaction with open-ended software,collaborative office assistants, and integrated development environments. Despite the prevalence of suchsettings in the real world, there is scarce work in formalizing the connection between high-level goals andlow-level behavior and inferring the former from the latter in these settings. We present a formal grammarfor describing users’ activities in such domains. We describe a new top-down plan recognition algorithmcalled CRADLE (Cumulative Recognition of Activities and Decreasing Load of Explanations) that uses thisgrammar to recognize agents’ interactions in exploratory domains. We compare the performance of CRADLEwith state-of-the-art plan recognition algorithms in several experimental settings consisting of real andsimulated data. Our results show that CRADLE was able to output plans exponentially more quickly thanthe state-of-the-art without compromising its correctness, as determined by domain experts. Our approachcan form the basis of future systems that use plan recognition to provide real-time support to users in agrowing class of interesting and challenging domains