Brown, S.M., Santos, E., Jr., and Banks, S.B., "Utility Theory-Based User Models for Intelligent Interface Agents", Proc. of the 12th Biennial Conf. of the Canadian Society for Computational Studies of Intelligence, 379-393, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....in the environment it situates, and actions to a ect changes in the environment. Here the problem is how to choose a proper action against the changes of the environment. Barbara It is interesting that an agent could be integrated into an expert system. For example, Brown, Santos and Banks [6] integrate their intelligent interface agents into an expert system shell called PESKI (Probabilities, Experts Systems, Knowledge, and Inference) 26, 27, 5] In [36] Knapik and Johnson further discuss how agents can make use of expert systems that are up and running in most domains in which ....

S.M. Brown, E.S. Jr., and S.B. Banks, Utility Theory-Based User Models for Intelligent Infer-face Agents, Advances in Arti cial Intelligence, R.E. Merler and E. Neufeld (Eds.), Lecture Notes in Arti cial Intelligence, 1418, Springer, pp. 378392, 1998.

No context found.

Brown, S.M., Santos, E., Jr., and Banks, S.B., "Utility Theory-Based User Models for Intelligent Interface Agents", Proc. of the 12th Biennial Conf. of the Canadian Society for Computational Studies of Intelligence, 379-393, 1998.

No context found.

Brown, S.M., Santos, E., Jr., and Banks, S.B., "Utility Theory-Based User Models for Intelligent Interface Agents", Proc. of the 12th Biennial Conf. of the Canadian Society for Computational Studies of Intelligence, 379-393, 1998.

....and has a complete, intuitive understanding of such actions. This paper describes the first application, a probabilistic expert system development environment (PESKI) to benefit from the active user interface approach. For two other applications that use this approach, see Brown et al. [3]. A suite of intelligent knowledge engineering tools (agents) have been developed and integrated using the active user interfaces paradigm. Simply put, our ultimate goal for PESKI is to guarantee that any and all actions taken by the expert and machine in building a decision support system is ....

....the PESKI environment. Additionally, a user profile is maintained on each user of PESKI so assistance may be custom tailored to individual users. The interface agent determines the how, when, what, and why of offering assistance to the user by inferencing over the user model and utility functions [3]. The agent acts as a rational decision maker on behalf of the user, using the maximum expected utility principle of decision theory to choose the goal with the maximum expected utility and suggests that goal. The agent is capable of offering assistance for such goals as which agent to use to ....

Brown, S.M., Santos Jr., E., and Banks, S.B., "Utility Theory-Based User Models for Intelligent Interface Agents", Proc. of the 12 Biennial Conf. of the Canadian Society for Computational Studies of Intelligence, 379-393, 1998.

....base. 3. Architecture of the interface agent Kavanah is different from the other information retrieval system in that the interface agent help generate the proactive query to search the knowledge base. The interface agent is constructed based on the Core Interface Agent (CIA) architecture [4]. The purpose of this architecture is to provide assistance to the user by maintaining an accurate model of the user s interaction with the target system environment. A user interacts with a target system (e.g typing a new query) is reported to CIA architecture as observations. The interface agent ....

....on the expected utility function. When the active user interface fails to meet its requirement (i.e. U aui requirement is falling below the threshold of utility requirement) we need to identify which requirements are not being met and attempt to correct the problem by updating the user model [4]. We take the approach of having the interface agent request help from a set of correction adaptation agents. Each correction adaptation agent offers a bid to the active user interface. The one that most likely improves the interface agent s requirement utility will win the bid and gets the ....

Brown, S. M., Santos, E. Jr., and Banks, S. B. 1998. Utility theory-based user models for intelligent interface agents. In Lecture Notes in Artificial Intelligence 1418: Advances in Artificial Intelligence -- AI '98, 378-392, Springer-Verlag.

....of people and institutions who are doing research on this disease. An accurate user model is considered necessary for effective ascription of user intent. We use utility theory and Bayesian network techniques to construct the user model on the fly from the user s interactions with the system [4], 13] By maintaining the dynamic user model, it allows the system to start with no knowledge of the user and incrementally build the user model as the user interacts with the system. Our interface agent is designed to support the complex notion of dynamic interests and preferences given the ....

....in the knowledge base to construct the proactive query on the user s behalf not only to retrieve all the relevant information for a particular query, but also to infer the user s overall goal in searching. 3. Framework The interface agent is based on the Core Interface Agent (CIA) architecture [4]. The purpose of this architecture is to provide assistance to the user by maintaining an accurate model of the user s interaction with the target system environment. A user interacts with a target system (e.g. a medical database querying system) typically via direct manipulation interface such ....

[Article contains additional citation context not shown here]

Brown, S. M., Santos, E. Jr., and Banks, S. B. 1998. Utility theory-based user models for intelligent interface agents. In Lecture Notes in Artificial Intelligence 1418: Advances in Artificial Intelligence -- AI '98, 378-392, Springer-Verlag.

.... [18] In recent years, a number of systems have used numerical uncertainty techniques from the AI community to capture the uncertainty inherent in modeling users [17] Our own research in the field of intelligent interface agents is demonstrated by our Core Interface Agent Architecture (CIaA) 2 [7], integrated into an expert system shell called PESKI [12, 13] PESKI (Probabilities, Expert Systems, Knowledge, and Inference) is an integrated probabilistic knowledge based expert system shell. PESKI provides users with knowledge acquisition [27] verification and validation [4, 28] data mining ....

....For more information on PESKI, see the Air Force Institute of Technology s Artificial Intelligence Laboratory web site 3 . CIaA supports effective user intent prediction by incorporating the ability to model both the uncertainty in user intent and dynamic user behavior within its user model [7]. To effectively predict user intent, an accurate cognitive model of the user is considered to be necessary. The problem with most cognitive models for intelligent interface agents is they rely on knowledge representations lacking flexibility and power in two key areas: the representation of ....

[Article contains additional citation context not shown here]

Scott M. Brown, Eugene Santos Jr., and Sheila B. Banks. Utility theory-based user models for intelligent interface agents. In Proceedings of the Twelfth Canadian Conference on Artificial Intelligence (AI '98), June 1998. to appear.

No context found.

S. M. Brown, E. S. Jr. and S. B. Banks, \Utility theory-based user models for intelligent interface agents", in Advances in Arti cial Intelligence, eds. R. E. Merler and E. Neufeld, LNAI 1118 (Springer, 1998), pp. 378-392.

No context found.

S. M. Brown, E. S. Jr. and S. B. Banks, "Utility theory-based user models for intelligent interface agents", in Advances in Artificial Intelligence, eds. R. E. Metlet and E. Neufeld, LNAI 1118 (Springer, 1998), pp. 378-392.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC