@inproceedings{Mohan2017b, author = {Shiwali Mohan and Anusha Venkatakrishnan and Danny Bobrow and Peter Pirolli}, title = {Health Behavior Coaching: A Motivating Domain for Human-Aware Artificial Intelligence Research}, booktitle = {In the Workshop on Human-Aware Artificial Intelligence, AAAI 2017}, year = {2017}, abstract = {Human-agent interaction has been studied for a while in AI research. However, very little attention has been paid to developing methods that are human-aware-that can model, reason about, and make decisions based on changes in human physiological, cognitive, and affec-tive states. We argue that health behavior coaching is a domain that requires an agent to be human-aware, in order to be effective and therefore, is a fruitful domain to pursue. We have delineated a set of desiderata for agents that can coach people to develop healthy behaviors. This set results from our ongoing work on developing AFYA (Mohan et al. 2017)-an interactive coach that resides in a smartphone app and coaches people exercise more. This paper describes each desiderata, the challenge it poses to AI research, and provides examples from our work demonstrating how it can be met.}, papertype = {cpaper}, pdf = {./content/health-behavior-coaching.pdf}, } @inproceedings{Mohan2017, author = {Shiwali Mohan and Anusha Venkatakrishnan and Michael Silva and Peter Pirolli}, title = {On Designing a Social Coach to Promote Regular Aerobic Exercise.}, booktitle = {Proceedings of the 29th Annual Conference on Innovative Applications of Artificial Intelligence/AAAI}, year = {2017}, abstract = {Our research aims at developing interactive, social agents that can coach people to learn new tasks, skills, and habits. In this paper, we focus on coaching sedentary, overweight individuals to exercise regularly. We employ adaptive goal setting in which the coach generates, tracks, and revises personalized exercise goals for a trainee. The goals become incrementally more difficult as the trainee progresses through the training program. Our approach is model-based - the coach maintains a parameterized model of the trainee’s aerobic capability that drives its expectation of the trainee’s performance. The model is continually revised based on interactions with the trainee. The coach is embodied in a smartphone application which serves as a medium for coach-trainee interaction. We show that our approach can adapt the trainee program not only to several trainees with different capabilities but also to how a trainee’s capability improves as they begin to exercise more. Experts rate the goals selected by the coach better than other plausible goals, demonstrating that our approach is effective.}, papertype = {cpaper}, pdf = {./content/designing-social-coach.pdf}, } @inproceedings{Li2016, author = {Justin Li and Steven Jones and Shiwali Mohan and Nate Derbinksy}, title = {Architectural Mechanisms for Mitigating Uncertainty during Long-Term Declarative Knowledge Access}, booktitle = {Proceedings of the 4th Conference on Advances in Cognitive Systems}, year = {2016}, abstract = {Long-term declarative semantic memory is a key component of a cognitive architecture that allows an agent to perform effectively in complex tasks. Over the years, a number of new mechanisms beyond deliberate cued retrieval have been developed, including partial matching and spreading activation. However, there has yet to be a theory that coherently explains why these mechanisms are beneficial or necessary. We propose that these semantic memory mechanisms are attempts to mitigate the uncertainty in the environment, as well as to compensate for the errors and incompleteness in agent knowledge. Through this lens, we show that existing mechanisms are misunderstood and motivate future areas of exploration, including how retrieval mechanisms should incorporate agent problem-solving context and how to unify the treatment of diverse mechanisms and the uncertainties they address.}, papertype = {cpaper}, url = {http://derbinsky.info/public/_custom/research/acs2016/paper.pdf}, } @inproceedings{Haztler2016, author={Andrea L. Hartzler* and Anusha Venkatakrishnan* and Shiwali Mohan and Michael Silva and Paula Lozano and James D. Ralston and Evette Ludman and Dori Rosenberg and Katherine and M. Newton and Lester Nelson and Peter Pirolli}, title={Acceptability of a team-based mobile health (mHealth) application for lifestyle self-management in individuals with chronic illnesses}, booktitle={Proceedings of the 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society}, year={2016}, abstract={With increased incidence of chronic illnesses arising due to unhealthy lifestyle habits, it is increasingly important to leverage technology applications to promote and sustain health behavior change. We developed a smartphone-based application, NutriWalking (NW), which recommends personalized daily exercise goals and promotes healthy nutritional habits in small peer teams. Here, we demonstrate an early study of usability and acceptability of this app in patients with type 2 Diabetes Mellitus and Depression. Our goal was to evaluate the potential of NW as a self-management support tool. Findings point to design considerations for team-based self-management tools delivered via mHealth platforms.}, papertype={cpaper}, url={http://ieeexplore.ieee.org/abstract/document/7591428/}, } @inproceedings{Pirolli2016, author = {Peter Pirolli and Shiwali Mohan and Rong Yang and Anusha Venkatakrishnan and Michael Silva and Michael Youngblood and Ashwin Ram and Les Nelson}, title = {User Modeling and Planning for Improving Self-efficacy and Goal Adherence in mHealth}, booktitle = {2nd Behaviour Change Conference: Digital Health and Wellbeing, London, United Kingdom}, year = {2016}, abstract = {mHealth applications provide great opportunities for projecting behavior-change methods into everyday life at large economies of scale. We have developed a smartphone-based system that provides support for health behavior change in diet, nutrition, and stress reduction. A key component of the system is an artificially intelligent (AI) coaching agent that personalizes the selection of behavioral goals (e.g., walking for 30 minutes). The aim of this personalization is to maximize daily goal adherence as well as long-run physical and psychosocial gains. We describe two modules that are central to this personalized coaching: (1) an AI planner that dynamically adjusts the schedule of future goals in reaction to a user’s compliance with the schedule and (2) a predictive cognitive model of self-efficacy based on neurocomputational theory that continuously changes in reaction to individual achievements or failures.}, papertype = {abstract}, url = {http://www.frontiersin.org/Community/AbstractDetails.aspx?ABS_DOI=10.3389/conf.FPUBH.2016.01.00107&eid=3118&sname=2nd_Behaviour_Change_Conference_Digital_Health_and_Wellbeing}, poster = {./content/cbc-poster.pdf}, } @inproceedings{Mohan2015, author = {Shiwali Mohan and James Kirk and Aaron Mininger and John Laird}, title = {Agent Requirements for Effective and Efficient Task-Oriented Dialog}, booktitle = {In the AAAI 2015 Fall Symposium Series}, year = {2015}, abstract = {Dialog is a useful way for a robotic agent performing a task to communicate with a human collaborator, as it is a rich source of information for both the agent and the human. Such task-oriented dialog provides a medium for commanding, informing, teaching, and correcting a robot. Robotic agents engaging in dialog must be able to interpret a wide variety of sentences and supplement the dialog with information from its context, history, learned knowledge, and from non-linguistic interactions. We have identified a set of nine system-level requirements for such agents that help them support more effective, efficient, and general taskoriented dialog. This set is inspired by our research in Interactive Task Learning with a robotic agent named Rosie. This paper defines each requirement and gives examples of work we have done that illustrates them.}, papertype = {wpaper}, pdf = {./content/mohan_aaai_fss_2015.pdf}, url = {http://www.aaai.org/ocs/index.php/FSS/FSS15/paper/view/11701}, talk = {./content/aaai-fss_talk.pdf}, } @inproceedings{Mohan2015, author = {Shiwali Mohan}, year = {2015}, title = {From Verbs to Tasks: An Integrated Account of Learning Tasks from Situated Interactive Instruction.}, booktitle = {Ph.D. Thesis, University of Michigan}, pdf={./content/mohan_thesis_2015.pdf}, abstract={Intelligent collaborative agents are becoming common in the human society. From virtual assistants such as Siri and Google Now to assistive robots, they contribute to human activities in a variety of ways. As they become more pervasive, the challenge of customizing them to a variety of environments and tasks becomes critical. It is infeasible for engineers to program them for each individual use. Our research aims at building interactive robots and agents that adapt to new environments autonomously by interacting with human users using natural modalities. This dissertation studies the problem of learning novel tasks from human-agent dialog. We propose a novel approach for interactive task learning, situated interactive instruction (SII), and investigate approaches to three computational challenges that arise in designing SII agents: situated comprehension, mixed-initiative interaction, and interactive task learning. We propose a novel mixed-modality grounded representation for task verbs which encompasses their lexical, semantic, and task-oriented aspects. This representation is useful in situated comprehension and can be learned through human-agent interactions. We introduce the Indexical Model of comprehension that can exploit extra-linguistic contexts for resolving semantic ambiguities in situated comprehension of task commands. The Indexical model is integrated with a mixed-initiative interaction model that facilitates a flexible task-oriented human-agent dialog. This dialog serves as the basis of interactive task learning. We propose an interactive variation of explanation-based learning that can acquire the proposed representation. We demonstrate that our learning paradigm is efficient, can transfer knowledge between structurally similar tasks, integrates agent-driven exploration with instructional learning, and can acquire several tasks. The methods proposed in this thesis are integrated in Rosie - a generally instructable agent developed in the Soar cognitive architecture and embodied on a table-top robot.}, papertype={treport}, } @inproceedings{Mohan2014aaai, author = {Shiwali Mohan and John E. Laird}, title = {Learning Goal-Oriented Hierarchical Tasks from Situated Interactive Instruction}, booktitle = {In the Proceedings of the 28th AAAI Conference on Artificial Intelligence}, year = {2014}, pdf={./content/mohan_AAAI_2014.pdf}, url={http://www.aaai.org/ocs/index.php/AAAI/AAAI14/paper/view/8630}, talk={}, poster={}, papertype={cpaper}, abstract={Our research aims at building interactive robots and agents that can expand their knowledge by interacting with human users. In this paper, we focus on learning goal-oriented tasks from situated interactive instructions. Learning the structure of novel tasks and how to execute them is a challenging computational problem requiring the agent to acquire a variety of knowledge including goal definitions and hierarchical control information. We frame acquisition of novel tasks as an explanation-based learning (EBL) problem and propose an interactive learning variant of EBL for a robotic agent. We show that our approach can exploit information in situated instructions along with the domain knowledge to demonstrate fast generalization on several tasks. The knowledge acquired transfers across structurally similar tasks. Finally, we show that our approach seamlessly combines agent-driven exploration with instructions for mixed-initiative learning.}, } @inproceedings{Mohan2014acs, author = {Shiwali Mohan and Aaron Mininger andJohn E. Laird}, title = {Towards an Indexical Model of Situated Comprehension for Cognitive Agents in Physical Worlds}, booktitle = {Advances in Cognitive Systems 3}, year = {2014}, pdf={./content/mohan_ACS_2014.pdf}, url={http://www.cogsys.org/pdf/paper-9-3-34.pdf}, papertype={jpaper}, abstract={We propose a computational model of situated language comprehension based on the Indexical Hypothesis that generates meaning representations by translating amodal linguistic symbols to modal representations of beliefs, knowledge, and experience external to the linguistic system. This Indexical Model incorporates multiple information sources, including perceptions, domain knowledge, and short-term and long-term experiences during comprehension. We show that exploiting diverse information sources can alleviate ambiguities that arise from contextual use of underspecific referring expressions and unexpressed argument alternations of verbs. The model is being used to support linguistic interactions in Rosie, an agent implemented in Soar that learns from instruction.}, } } @inproceedings{Laird2014bica, author = {John E. Laird and Shiwali Mohan}, title = {A Case Study of Knowledge Integration Across Multiple Memories in Soar}, booktitle = {Biologically Inspired Cognitive Architectures (invited)}, year = {2014}, pdf = {./content/laird_BICA2014.pdf}, url = {http://www.sciencedirect.com/science/article/pii/S2212683X14000164}, papertype = {jpaper}, abstract = {Online perception, behavior, and learning in complex domains require an intelligent agent to quickly and reliably access different types of knowledge. A cognitive architecture, therefore, must implement a diverse set of memories that are optimized for storing, accessing, and learning these different types of knowledge. In this paper, we describe a complex Soar agent that uses and learns multiple types of knowledge while interacting with a human in a real-world domain. Our hypothesis is that a diverse set of memories is required for the different types of knowledge. We first present the agent’s processing, highlighting the types of knowledge used for each phase. We then present Soar’s memories and identify which memory is used for each type of knowledge. We also analyze which properties of each memory make it appropriate for the knowledge it encodes. We conclude with a summary of our analysis. }, } @inproceedings{Mohan2014a, author = {Shiwali Mohan and John E. Laird}, title = {Learning New Tasks from Situated Interactive Instruction}, booktitle = {In the 2014 HRI Pioneers workshop}, year = {2014}, papertype = {wpaper}, pdf = {./content/mohan_HRIPioneers_2014.pdf}, url = {http://www.hripioneers.info/Proceedings/2014PioneersProceedings.pdf}, poster = {./content/mohan_HRIPioneers_2014_poster.pdf},} @inproceedings{Mohan2013acs, author = {Shiwali Mohan and Aaron Mininger and John E. Laird}, title = {Towards an Indexical Model of Situated Language Comprehension for Real-World Cognitive Agents}, booktitle = {In the Second Annual Conference on Advances in Cognitive Systems}, year = {2013}, papertype = {cpaper}, url={http://www.cogsys.org/papers/2013conference29.pdf},} @inproceedings{Laird2013, author = {John E. Laird and Shiwali Mohan}, title = {A Case Study of Knowledge Integration across Multiple Memories in Soar}, booktitle = {In Papers from Integrated Cognition (AAAI Fall Symposium Series)}, year = {2013}, papertype = {wpaper}, pdf = {./content/laird_AAAI_IC_2013.pdf}, url = {http://www.aaai.org/ocs/index.php/FSS/FSS13/paper/view/7606}, } @inproceedings{Mohan2012f, author = {Shiwali Mohan and James Kirk and John Laird}, title = {A Computational Model of Situated Task Learning with Interactive Instruction}, booktitle = {In Proceedings of the 17th International Conference on Cognitive Modeling}, year = {2013}, pdf = {./content/mohan_ICCM_2013.pdf}, talk = {./content/mohan-iccm-talk.pdf}, url = {http://iccm-conference.org/2013-proceedings/papers/0049/index.html}, papertype = {cpaper}, } @inproceedings{Mohan2012f, author = {Shiwali Mohan and Aaron Mininger and James Kirk and John Laird}, title = {Acquiring Grounded Representations of Words with Situated Interactive Instruction}, booktitle = {Advances in Cognitive Systems, 2}, year = {2012}, pdf = {./content/mohan_ACS_2012.pdf}, papertype = {jpaper}, url = {http://www.cogsys.org/pdf/paper-3-2-136.pdf}, talk = {./content/acs-talk.pdf}, } @inproceedings{Joshi2012a, author = {Mandar Joshi and Rakesh Khobragade and Saurabh Sarda and Umesh Deshpande and Shiwali Mohan}, title = {Object-Oriented Representation and Hierarchical Reinforcement Learning in Infinite Mario}, booktitle = {In Proceedings of the 24th IEEE International Conference on Tools with Artificial Intelligence}, year = {2012}, pdf = {./content/joshi_ICTAI_2012.pdf}, url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6495169}, papertype = {cpaper},} @inproceedings{Mohan2012e, author = {Shiwali Mohan* and Aaron Mininger* and James Kirk* and John Laird}, title = {Learning Grounded Language through Situated Interactive Instruction}, booktitle = {In Papers from Robots Learning Interactively from Human Teachers (AAAI Fall Symposium Series)}, pdf = {./content/mohan_AAAIFS_2012.pdf}, url = {http://aaai.org/ocs/index.php/FSS/FSS12/paper/view/5662}, year = {2012}, papertype = {wpaper}, url = {http://www.aaai.org/ocs/index.php/FSS/FSS12/paper/view/5662}, talk = {./content/aaaifs-talk.pdf}, } @inproceedings{Joshi2012, author = {Mandar Joshi and Rakesh Khobragade and Saurabh Sarda and Umesh Deshpande and Shiwali Mohan}, title = {Hierarchical Action Selection for Reinforcement Learning in Infinite Mario}, booktitle = {In Proceedings of the 6th Starting Artificial Intelligence Research Symposium (co-located with ECAI)}, year = {2012}, pdf = {./content/joshi_STAIRS_2012.pdf}, url = {http://plata.ar.media.kyoto-u.ac.jp/mori/research/Proceedings/ECAI2012/content/stairs/stairs201215.pdf}, papertype = {wpaper}, url = {http://books.google.com/books?hl=en&lr=&id=WOc8WSwcCjoC&oi=fnd&pg=PA162&dq=info:Zp20TtDieTIJ:scholar.google.com&ots=u-dG_96A95&sig=X1HmRu-UJj4UZ-8Y2n3YU-SO_eI}, } @inproceedings{Mohan2012d, author = {John Laird and Keegan Kinkade and Shiwali Mohan and Joseph Xu}, title = {Cognitive Robotics Using the Soar Cognitive Architecture}, booktitle = {In Proceedings of the 8th International Cognitive Robotics Workshop}, year = {2012}, pdf = {./content/laird_AAAICogRob_2012.pdf}, url ={http://aaai.org/ocs/index.php/WS/AAAIW12/paper/view/5221}, papertype = {wpaper}, } @inproceedings{Mohan2012c, author = {Shiwali Mohan and John Laird}, title = {Situated Comprehension of Imperative Sentences in Embodied, Cognitive Agents}, booktitle = {Grounding Language for Physical Systems, AAAI Technical Report WS-12-07}, year = {2012}, pdf = {./content/mohan_AAAIGPS_2012.pdf}, url = {http://aaai.org/ocs/index.php/WS/AAAIW12/paper/view/5245}, papertype= {wpaper}, } @inproceedings{Mohan2012b, author = {Shiwali Mohan and John Laird}, title = {Exploring Mixed-Initiative Interaction for Learning with Situated Instruction in Cognitive Agents}, booktitle = {Proceedings of the 26th AAAI Conference on Artificial Intelligence}, year = {2012}, pdf = {./content/mohan_AAAISA_2012.pdf}, url = {http://www.aaai.org/ocs/index.php/AAAI/AAAI12/paper/view/4834}, papertype = {abstract}, note = {\textit{(Extended Abstract)}}, } @inproceedings{Mohan2012a, author = {Shiwali Mohan and John Laird}, title = {Learning Actions and Action Verbs from Human-Agent Interaction}, booktitle = {17th AAAI/SIGART Doctoral Consortium}, year = {2012}, pdf = {./content/mohan_AAAIDC_2012.pdf}, papertype = {abstract}, url = {http://www.aaai.org/ocs/index.php/AAAI/AAAI12/paper/viewFile/4856/5288}, note = {\textit{(Extended Abstract)}}, talk = {./content/dc-r.pdf}, } @inproceedings{Mohan2011a, author = {Shiwali Mohan and John Laird}, title = {Towards Situated, Interactive, Instructable Agents in a Cognitive Architecture}, booktitle = {Papers from the 2011 AAAI Fall Symposium Series}, year = {2011}, keywords = {cognition; Soar; learning with instruction; human agent collaboration; rule-based systems}, abstract = {This paper discusses the challenge of designing instructable agents that can learn through interaction with a human expert. Learning through instruction is a powerful paradigm for acquiring knowledge because it limits the complexity of the learning task in a variety of ways. To support learning through instruction, the agent must be able to effectively communicate its lack of knowledge to the human, comprehend instructions, and apply them to the ongoing task. Weidentify some problems of concern when designing instructable agents. We propose an agent design that addresses some of these problems. We instantiate this design in the Soar cognitive architecture and analyze its capabilities on a learning task.}, url = {http://www.aaai.org/ocs/index.php/FSS/FSS11/paper/view/4165}, pdf = {./content/mohan_fss_2011.pdf}, papertype = {cpaper}, } @inproceedings{Mohan2011b, author = {Shiwali Mohan and John Laird}, title = {An Object-Oriented Approach to Reinforcement Learning in an Action Game}, booktitle = {Proceedings of the Artificial Intelligence for Interactive Digital Entertainment Conference}, keywords = {decision making; reinforcement learning; action games}, abstract = {In this work, we look at the challenge of learning in an action game,Infinite Mario. Learning to play an action game can be divided into two distinct but related problems, learning an object-related behavior and selecting a primitive action. We propose a framework that allows for the use of reinforcement learning for both of these problems. We present promising results in some instances of the game and identify some problems that might affect learning.}, url = {http://www.aaai.org/ocs/index.php/AIIDE/AIIDE11/paper/view/4069}, series = {AIIDE}, year = {2011}, pdf = {./content/mohan_aiide_2011.pdf}, papertype = {cpaper}, } @inproceedings{Mohan2010, author = {Shiwali Mohan and John Laird}, title = {Relational Reinforcement Learning in Infinite Mario}, booktitle = {Proceedings of the 24th AAAI Conference on Artificial Intelligence}, abstract = {Relational representations in reinforcement learning allow for the use of structural information like the presence of objects and relationships between them in the description of value functions. Through this paper, we show that such representations allow for the inclusion of background knowledge that qualitatively describes a state and can be used to design agents that demonstrate learning behavior in domains with large state and actions spaces such as computer games.`}, year = {2010}, url = {http://www.aaai.org/ocs/index.php/AAAI/AAAI10/paper/view/1657}, pdf = {./content/mohan.pdf}, note = {\textit{(Extended Abstract)}}, papertype = {abstract}, } @techreport{Mohan2009, author = {Shiwali Mohan and John E. Laird}, title = {Learning to Play Mario}, NUMBER = {CCA-TR-2009-03}, booktitle = {Technical Report CCA-TR-2009-03 Center for Cognitive Architecture, University of Michigan, Ann Arbor, Michigan}, INSTITUTION = {Center for Cognitive Architecture, University of Michigan}, ADDRESS = {Ann Arbor, Michigan}, ABSTRACT = {Computer Games are interesting test beds for research in Artificial Intelligence and Machine Learning. Games usually have continuous state spaces, large action spaces and are characterized by complex relationships between components. Without applying abstraction and generalizations, learning in computer games domain becomes infeasible. Through this work, we investigate some designs that facilitate tractable reinforcement learning in symbolic agents developed using Soar architecture operating in a complex domain, Infinite Mario. Object oriented representations of the environment greatly simplify otherwise complex state spaces. We also demonstrate that imposing hierarchy in problem structure greatly reduces the complexity of the tasks and aids in learning generalized policies that can be transferred across similar tasks.}, year = {2009}, url = {http://sitemaker.umich.edu/SoarWeb/Publications/da.data/000000000000000000000000000000000000000003005536/Downloadpaper/filename}, papertype = {treport}, pdf = {http://sitemaker.umich.edu/SoarWeb/Publications/da.data/000000000000000000000000000000000000000003005536/Downloadpaper/filename}, } @inproceedings{Mohan2008, author = {Niladri Chatterjee and Shiwali Mohan}, title = {Discovering Word Senses from Text using Random Indexing}, booktitle = {Proceedings of the 9th International Conference on Computational linguistics and Intelligent Text Processing}, abstract = {Random Indexing is a novel technique for dimensionality reduction while creating Word Space model from a given text. This paper explores the possible application of Random Indexing in discovering word senses from the text. The words appearing in the text are plotted onto a multi-dimensional Word Space using Random Indexing. The geometric distance between words is used as an indicative of their semantic similarity. Soft Clustering by Committee algorithm (CBC) has been used to constellate similar words. The present work shows that the Word Space model can be used effectively to determine the similarity index required for clustering. The approach does not require parsers, lexicons or any other resources which are traditionally used in sense disambiguation of words. The proposed approach has been applied to TASA corpus and encouraging results have been obtained.}, series = {CICLing}, year = {2008}, note = {\textbf{Best Paper Award}}, url = {http://www.springerlink.com/content/xp70kw14w0054541/}, pdf = {./content/mohan_cicling_2008.pdf}, papertype = {cpaper}, } @inproceedings{Mohan2007, author = {Niladri Chatterjee and Shiwali Mohan}, title = {Extraction-Based Single-Document Summarization Using Random Indexing}, booktitle ={Proceeding of the 19th IEEE International Conference on Tools with Artificial Intelligence}, abstract = {This paper presents a summarization technique for text documents exploiting the semantic similarity between sentences to remove the redundancy from the text. Semantic similarity scores are computed by mapping the sentences on a semantic space using Random Indexing. Random Indexing, in comparison with other semantic space algorithms, presents a computationally efficient way of implicit dimensionality reduction. It involves inexpensive vector computations such as addition. It thus provides an efficient way to compute similarities between words, sentences and documents. Random Indexing has been used to compute the semantic similarity scores of sentences and graph-based ranking algorithms have been employed to produce an extract of the given text.}, year = {2007}, url ={http://www.computer.org/portal/web/csdl/doi/10.1109/ICTAI.2007.28}, pdf ={./content/mohan_ictai.pdf}, papertype = {cpaper}, }