W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The Distributed Agent Architecture Of The University of Michigan Digital Library (UMDL). In AAAI Spring Symposium Series on Software Agents, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....collaboration is an explicit action, and the rules governing such collaborations can be encoded as rules within agent programs. Agent Architectures. For an excellent anthology of classic works on agent systems, see [49] There have been numerous proposals for agentization in the literature (e.g. [35, 43, 17]) which have been broadly classified by Genesereth and Ketchpel [40] into four categories: in the first category, each agent has an associated transducer that converts all incoming messages and requests into a form that is intelligible to the agent. This is clearly not what happens in IMPACT ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The Distributed Agent Architecture Of The University of Michigan Digital Library (UMDL). In AAAI Spring Symposium Series on Software Agents, 1995.

....discovery from databases approach information flood problem from information sources side [12, 9] In these studies, implemented systems try to extract data ordinalities or relativities. Recently numerous studies have been made on mediating between information sources. Agent based approach [4, 20], integration of schemes of databases [14] are currently hot issues. Studies on HCI for IR approach from the users side. Visualization of information space helps users to recognize features of information or relations among objects. Many studies focus on visualization from geometrical aspects ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the university of michigan digital library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, pages 19--24, 1995.

....discovery from databases approach information flood problem from information sources side [12, 9] In these studies, implemented systems try to extract data ordinalities or relativities. Recently numerous studies have been made on mediating between information sources. Agentbased approach [4, 22], integration of schemes of databases [14] are currently hot issues. Studies on HCI for IR approach from the users side. Visualization of information space helps users to recognize features of information or relations among objects. Many studies focus on visualization from geometrical aspects ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the university of michigan digital library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, pages 19--24, 1995.

....Mining and Knowledge Discovery from Databases approach this problem from information sources side [1,2] In these studies, implemented systems try to extract data ordinalities or relativities. Recently numerous studies have been made on mediating between information sources. Agent based approach [3,4], integration of schemes of databases [5] are currently hot issues. Studies on visualization of information space, on the other hand, approach from the users side. We pointed out that it is effective that an interactive system lets its user be conscious of concepts and their relations [6] We, ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman, "The Distributed Agent Architecture of the University of Michigan Digital Library, " AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, 1995, pp.19 -- 24.

.... global properties may emerge from the collective interactions of individual agents [Kurose et al. 1985, Huberman, 1988, Clearwater, 1995] Much of the latter work falls under the rubric of market based control , in which economic transactions are used to bring about some predefined, desired end [Birmingham et al. 1996, Wellman, 1993, Stonebraker and others, 1994, Clearwater, 1995] Agents may be designed to cooperate [Huberman et al. 1996] or to compete [Hogg and Huberman, 1991] but so long as the aggregate evolves 1 toward a globally defined optimum, the system as a whole is deemed successful. But in an ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

....data. Preliminary evaluation shows that when operating under resource constraints (related to cost of information access and limited amount of time) our information gathering strategy leads to significant performance improvements. Although much work has been done on information gathering [1, 3, 9] and decision making [6, 7, 4] separately, little work has capitalized on the synergy that can develop when these two problems are solved together [8, 11, 13] Previous work on using information value theory has concentrated on a small set of information sources with little uncertainty regarding ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the university of michigan digital library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, pages 19--24, Stanford, CA, 1995.

....defines messages as performatives. AgentTalk [8] supports inheritance mechanism using which a designer can introduce new protocols based on existing protocols. Some of the popular agent systems include UNIX softbots [3] Indexing agents [4] FAQ finders [4] Internet softbots [4] and UMDL agents [1]. 3 Personal Assistant System : A Multi Agent Architecture The multi agent system architecture for the Personal Assistant system consists of an User Interface Agent, a Control Agent, a Name Server Agent, a Case Base Interface Agent and a number of Service Agents. The Service Agents (SA) embody ....

W. P. Birmingham et al. The Distributed Agent Architecture of the University of Michigan Digital Library, May 1996.

.... global properties may emerge from the collective interactions of individual agents (Kurose et al. 1985; Huberman, 1988; Clearwater, 1995) Much of the latter work falls under the rubric of market based control , in which economic transactions are used to bring about some predefined, desired end (Birmingham et al. 1996; Wellman, 1993; Stonebraker and others, 1994; Clearwater, 1995) Agents may be designed to cooperate (Huberman et al. 1996) or to compete (Hogg and Huberman, 1991) but so long as the aggregate evolves toward a globally defined optimum, the system as a whole is deemed successful. But in an open ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

....collaboration is an explicit action, and the rules governing such collaborations can be encoded as rules within agent programs. Agent Architectures. For an excellent anthology of classic works on agent systems, see [49] There have been numerous proposals for agentization in the literature (e.g. [35, 43, 17]) which have been broadly classified by Genesereth and Ketchpel [40] into four categories: in the first category, each agent has an associated transducer that converts all incoming messages and requests into a form that is intelligible to the agent. This is clearly not what happens in IMPACT ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The Distributed Agent Architecture Of The University of Michigan Digital Library (UMDL). In AAAI Spring Symposium Series on Software Agents, 1995.

....collaboration is an explicit action, and the rules governing such collaborations can be encoded as rules within agent programs. Agent Architectures. For an excellent anthology of classic works on agent systems, see [57] There have been numerous proposals for agentization in the literature (e.g. [38, 46, 16]) which have been broadly classified by Genesereth and Ketchpel [43] into four categories: in the first category, each agent has an associated transducer that converts all incoming messages and requests into a form that is intelligible to the agent. This is clearly not what happens in IMPACT ....

W. P. Birmingham, E. H. Durfee, T. Mullen and M. P. Wellman. The Distributed Agent Architecture Of The University of Michigan Digital Library (UMDL). In: Proc. AAAI Spring Symposium Series on Software Agent, 1995.

....themselves [2, 4, 12, 14] Is the emergence and survival of this economy guaranteed The answer is far from clear. Within the context of market driven optimization, a growing body of work demonstrates the success of economic mechanisms in coordinating the actions of agent populations [1, 3, 5, 11, 13]. But by definition, marketdriven optimization presupposes a common overall goal to be striven for, in the form of a predefined measure of system performance, along with the implicit power of the system s designers to tune its parameters to help it achieve its optimum. In an open system like the ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

....are available in house and if not how long it will take to get them and what they will cost [28] We see this automation of a sophisticated librarian as a natural step in the evolutionary development of a fully automated digital library. Our approach builds on current document location technology [1, 6, 15, 31, 32] by introducing the value of information and its cost, time and likelihood of being acquired as the driving force behind the decision of when, where and how to locate specific documents. 2 Overview of the System Our proposed system architecture is based on three primary layers that operate ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The Distributed Agent Architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, Stanford, CA, 1995.

.... disastrous collective behavior[13, 14] On the other hand, a large body of work on market mechanisms in distributed multi agent environments suggests that efficient resource allocation or other desirable global properties may emerge from the collective interactions of individual agents[1, 3, 5, 8, 10, 11, 15, 17, 21]. Our goal is to understand the dynamic, emergent behaviors both good and bad of information economies from an agent s eye view, and from this to formulate basic design principles that will foster efficient electronic commerce. We pursue this goal by combining analysis and simulation of ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

.... disastrous collective behavior[11, 12] On the other hand, a large body of work on market mechanisms in distributed multi agent environments suggests that efficient resource allocation or other desirable global properties may emerge from the collective interactions of individual agents[1, 3, 5, 7, 8, 9, 13, 14, 17]. We seek to understand the dynamic, emergent behaviors both good and bad of information economies from an agent s eye view, and from this to formulate basic design principles that will foster efficient electronic commerce. We pursue this goal by combining analysis and simulation of ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan Digital Library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

.... efficient resource allocation or other desirable global properties may emerge from the collective interactions of individual agents[8, 6, 3] Much of this work falls under the rubric of market based control , in which economic transactions are used to bring about some predefined, desired end[1, 15, 12, 3]. Agents may be designed to cooperate[7] or to compete[5] but so long as the aggregate evolves toward a globally defined optimum, the system as a whole is successful. But in an open system like the Web, there is no global purpose being served by the collective of agents; strictly speaking, there ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the university of michigan digital library. In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments. AAAI Press, 1996.

....source. Currently, the system manages the mapping between the type of information source and its name and address by simple table lookups. In the next phase, sophisticated mediators [ Wiederhold, 1992 ] will provide a more flexible mapping, similar to the mechanisms used in the UMDL project [ Birmingham, 1995; Birmingham et al. 1995 ] There, so called registry agents capture the address and content information of each database. Query planning agents negotiate with the registry agents about particular addresses of databases (see Fig. 3) During the negotiation phase, facilitator agents might be ....

....the system manages the mapping between the type of information source and its name and address by simple table lookups. In the next phase, sophisticated mediators [ Wiederhold, 1992 ] will provide a more flexible mapping, similar to the mechanisms used in the UMDL project [ Birmingham, 1995; Birmingham et al. 1995 ] There, so called registry agents capture the address and content information of each database. Query planning agents negotiate with the registry agents about particular addresses of databases (see Fig. 3) During the negotiation phase, facilitator agents might be involved to resolve different ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The distributed agent architecture of the University of Michigan digital library. In Proc. AAAI Spring Symp. Series on Information Gathering from Distributed Heterogeneous Environments, Stanford, CA, March 1995

....constraints The existing Distributed Information Retrieval systems have a significant influence in the architectural as well as the implementation features of our system. The concept of some of the agent classes with a role similar to ours exists in UMDL (such as the collection interface agents) Birmingham 95] and [Vidal 95] IRA (such as userbots, corpusbots) Voorhees 94] TSIMMIS (such as classifiers, translators) Garcia Molina 95] Information Brokers [Fikes 95] SHADE COINS [Kuokka 95] and Knowledge Navigator (such as advisory agents) Burke 95] among others. Other research work on multimedia ....

Birmingham, W. P., Durfee, E.H. The Distributed Agent Architecture of the University of Michigan Library. In AAAI95 Symposium: Information Gathering from Heterogeneous, Distributed Environments.

.... [13, 16] adds new dimensions to traditional document based information retrieval[11] Vast amounts of heterogeneous information are distributed across local and remote locations[3] The agent paradigm provides a metaphorical bridge that permits transparency of access to this information[17, 5, 1, 2, 14]. Both the distributed nature of the data and agent based access also broaden the notion of relevance[15] Relevance can acquire new dimensions, e.g. of timeliness, comprehensiveness, and authoritativeness of the retrievables. However, all of these increase the information load which must be ....

W. P. Birmingham, E. H. Durfee, T. Mullen, and M. P. Wellman. The Distributed Agent Architecture of the University of Michigan Digital Library. University of Michigan Department of Computer Science Technical Memo. (1995)

....Distributed Information Retrieval systems have a significant influence in the architectural as well as the implementation features of our system. The concept of some of the agent classes with a role similar to ours exists (among others) in: a) UMDL (such as the collection interface agents) Birmingham 95] and [Vidal 95] b) IRA (such as userbots, corpusbots) Voorhees 94] c) TSIMMIS (such as classifiers, translators) Garcia Molina 95] d) Information Brokers [Fikes 95] e) SHADE COINS [Kuokka 95] and f) Knowledge Navigator (such as advisory agents) Burke 95] All the above systems tend to ....

Birmingham, W. P., Durfee, E.H. 1995. The Distributed Agent Architecture of the University of Michigan Digital Library. In AAAI 95: Information Gathering from Heterogeneous, Distributed Environments.

....developed programs available on the Web, and this number is increasing on a regular basis. However, we do not address issues such as load balancing and privacy issues addressed by Decker et al. 3] With respect to agent architectures, there have been numerous proposals in the literature (e.g. [4, 6, 1]) which have been broadly classified by Genesereth and Ketchpel[5] into four categories: in the first category, each agent has an associated transducer that converts all incoming messages and requests into a form that is intelligible to the agent. This is clearly not what happens in IMPACT ....

W. P. Birmingham, E. H. Durfee, T. Mullen and M. P. Wellman. (1995) The Distributed Agent Architecture Of The University of Michigan Digital Library (UMDL), Spring Symposium Series on Software Agent.

....is the primary activity. Consumers in the computational economy correspond to end users of information services, and producers correspond to companies, individuals, or software agents supplying these services. An example of such an environment is the University of Michigan Digital Library, UMDL) [2, 3], which aims to provide end users with access to a multitude of highly distributed and constantly changing information sources and services. The design of UMDL is based on a distributed agent architecture, depicted in Figure 1. The users in this figure users collections mediators Fig. 1. UMDL ....

William P. Birmingham, Edmund H. Durfee, Tracy Mullen, and Michael P. Wellman. The distributed agent architecture of the University of Michigan Digital Library (extended abstract) . In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments, Stanford, CA, March 1995.

.... programming to one large scale information services network, the University of Michigan Digital Library (Birmingham et al. 1994) Our design for this system is based on a network of specialized information agents, interacting as suppliers and producers in a virtual information services economy (Birmingham et al. 1995). Within this system, efficient allocation of basic computational resources (memory, processing, bandwidth) as well as information goods and more complex value added services will be a key to effective behavior of the overall system. Blue Skies Economy For our initial exploration of resource ....

Birmingham, W. P.; Durfee, E. H.; Mullen, T.; and Wellman, M. P. 1995. The distributed agent architecture of the University of Michigan Digital Library (extended abstract). In AAAI Spring Symposium on Information Gathering in Heterogeneous, Distributed Environments.

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