ANTIPA: an agent architecture for intelligent information assistance

Human users trying to plan and accomplish informationdependent goals in highly dynamic environments with prevalent uncertainty must consult various types of information sources in their decision-making processes while the information requirements change as they plan and re-plan. When the users must make timecritical decisions in information-intensive tasks they become cognitively overloaded not only by the planning activities but also by the information-gathering activities at various points in the planning process. We have developed the ANTicipatory Information and Planning Agent (ANTIPA) to manage information adaptively in order to mitigate user cognitive overload. To this end, the agent brings information to the user as a result of user requests but most crucially, it proactively predicts the user’s prospective information needs by recognizing the user’s plan; pre-fetches information that is likely to be used in the future; and offers the information when it is relevant to the current or future planning decisions. This paper introduces a fully implemented agent of the ANTIPA architecture using a decision-theoretic user model. 1 Information-dependent planning problem We focus on a class of problems where a user (or a planner) must access various types of information sources to acquire current information that is required for executing certain actions. Here, in addition to domain-specific planning objectives, the user must also take the cost of getting information into consideration in selecting actions. Furthermore, the quality of information (that also depends on the source of information) affects the user’s transition to another state after taking the action. For instance, consider a student preparing for a final exam by reviewing selected topics covered in a semester. When a question is encountered, the student may search online for a quick answer by taking a risk that the answer found may be incorrect, or email the teacher and wait to get a generally more precise answer. The outcome of the student’s action (to understand the concept) depends on the quality of information, which in turn depends on the source from which it has come. For example, by receiving high-quality information, the student’s state regarding the understanding of a certain concept is more likely to transition from not-learned to learned, 1 Research was sponsored by the Army Research Laboratory and was accomplished under Cooperative Agreement Number W911NF-09-2-0053. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation here on. 2 Robotics Institute, Carnegie Mellon University, USA, email: {jeanoh, meneguzzi, katia}@cs.cmu.edu also increasing the chance of getting a better grade (reward) in the final exam. Given that the user is trying to solve an information-dependent planning problem, we aim to develop an agent that can adaptively identify and manage the user’s information needs to facilitate the user’s actions.