E. Durfee, V. Lesser, and D. Corkill, "Trends in cooperative distributed problem solving," IEEE Transactions on Knowledge and Data Engineering, vol. KDE-1, no. 1, pp. 63--83, March 1989.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....environments or applications based on the same standard thus avoiding costly and time consuming integration procedures. 4 BDI Agents 4. 1 Belief Desire Intention (BDI) Much research work has been commissioned on Multi Agent Systems (MAS) and Distributed Artificial Intelligence (DAI) BG88] [DLC89], OJ96] Significant quantities of this research has been concerned with developing practical deliberative reasoning models. Perhaps the most successful attempt at constructing such models have come from the application of a mental state comprising a set of mental attitudes such as beliefs, ....

Durfee, E.H., Lesser, V.R., Corkhill, D.D., Trends in co-operative distributed problem solving, IEEE: Knowl. Data Eng. 11(1), 63-8, 1989.

....sense that they have specific abilities. The second subarea is represented by Distributed Problem Solving (DPS) which concerns homogeneous systems, which are systems where every agent may perform the 3 task of any other agent. A subclass of DPS is Cooperative Distributed Problem Solving (CDPS, see [13] for a survey of CDPS systems) a CDPS system is a network of problem solvers (i.e. systems of artificial intelligence) that work together. Among the many different models of agents that have been developed we can distinguish models which are utilized to describe agents at an implementational ....

E. H. Durfee, V. R. Lesser, D. D. Corkill, "Trends in Cooperative Distributed Problem Solving", IEEE Transactions on Knowledge and Data Engineering, vol. 1, no. 1, March 1989, p. 63-83

....objectives that are typically beyond the capabilities of any individual agent. In short, CDM has generally been viewed and modelled as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38]. Recently, a number of theoretical models have been proposed for investigating decisionmaking in a social setting [33, 48, 71, 82, 86] However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these ....

.... as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38] Recently, a number of theoretical models have been proposed for investigating decisionmaking in a social setting [33, 48, 71, 82, 86]. However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these approaches, agents are seen as endowed with identifiable decision making capabilities and are grouped together to form communities which ....

E. H. Durfee, V. R. Lesser and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Trans. On knowledge and Data Engineering, 1, 63--83, 1989.

....objectives that are typically beyond the capabilities of any individual agent. In short, CDM has generally been viewed and modelled as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38]. Recently, a number of theoretical models have been proposed for investigating decision making in a social setting [33, 48, 71, 82, 86] However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these ....

.... as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38] Recently, a number of theoretical models have been proposed for investigating decision making in a social setting [33, 48, 71, 82, 86]. However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these approaches, agents are seen as endowed with identifiable decision making capabilities and are grouped together to form communities which ....

E.H. Durfee, V. R. Lesser and D. D. Corkill. Trends in cooperative distributed problem solving. In IEEE Trans. On knowledge and Data Engineering, 1(1), pages 63-83, 1989.

....objectives that are typically beyond the capabilities of any individual agent. In short, CDM has generally been viewed and modelled as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38]. Recently, a number of theoretical models have been proposed for investigating decisionmaking in a social setting [33, 48, 71, 82, 86] However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these ....

.... as a kind of distributed reasoning and search, whereby a collection of agents collaboratively go through the search space of a problem in order to find a solution [11, 33, 38] Recently, a number of theoretical models have been proposed for investigating decisionmaking in a social setting [33, 48, 71, 82, 86]. However, none of these approaches cover the full breadth of social and cognitive activities that are typically involved in a CDM process. In some of these approaches, agents are seen as endowed with identifiable decision making capabilities and are grouped together to form communities which ....

E. H. Durfee, V. R. Lesser and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Trans. On knowledge and Data Engineering, 1, 63--83, 1989.

....but also prove that our algorithm is better than Katz and Rosenschein s algorithm both on time complexity and space complexity. Keywords: Planning, Verification, Multiple agent systems. 1 Introduction Distributed problem solving plays an important role in distributed artificial intelligence [2,10,17,19,20,21]. Now, it is fashionable to use planning as a kind of approach for distributed problem solving in multiple agent environments [4,7] Planning research in multiple agent systems has historically focussed on two distinct classes of problems. One paradigm has been that of planning for multiple ....

E. H. Durfee, V. R. Lesser and D. D. Corkill, Trends in cooperative distributed problem solving, IEEE Trans, Knowl. Data Eng. ,1989, 1: 63-83.

....42] Contributions come from a variety of sources. Work from Computing Science, Cognitive Science, Psychology and Social Science has addressed questions such as how best to characterise and implement collaborative interaction facilities and supporting tools for the cooperating end user community [16, 12, 13, 24, 44]. For example, Rodden [34] advocated a support layer running over distributed technology that would provide a separation between policies and sets of implementation mechanisms. Reddy [32] envisaged a layered architecture of technologies in order to support virtual teams. A virtual team is defined ....

Durfee, E.H., Lesser, V.R., Corkill, D.D., "Trends in Cooperative Distributed Problem Solving", IEEE Transactions on Knowledge and Data Engineering Vol. 1 No. 1:63-83, 1989.

....about changing is made every time the enterprise satisfaction level decreases more than a certain value, or after a certain period since the last decision. This process is very costly, both in communication and computation, and should be tuned so that reorganization is undertaken infrequently [3]. 2.2 Negotiation Protocols Due to the behavior duality, two communication protocols are used. Inside the enterprise, the allocation agents cooperate trough a slight variation of the original Contract Net protocol (CNP) 11] For negotiation processes between different enterprises it is used ....

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63-- 83, March 1989.

....can synthesize all uncertainty information encoded by probability functions or mass functions. 1 Introduction A distributed expert system (DES) consists of multiple physically separated processing nodes which have at least one expert system (ES) Some work has been done on cooperation in DES [1,4]. Important approaches of cooperation include negotiation, FA C, and organizational structuring. In [8] Zhang classifies the cooperation among component ESs in a DES into four types: horizontal cooperation, hierarchical cooperation, recursive cooperation and hybrid cooperation. In this paper we ....

E. H. Durfee, V. R. Lesser and D. D. Corkill, Trends in cooperative distributed problem solving, IEEE Trans, Knowl. Data Eng. ,1989, 1: 63-83.

....structure is made every time the enterprise satisfaction decreases more than a parameterized value, or after a certain period is passed since the last decision. This process is very costly, both in communication and computation, and should be tuned so that reorganization is undertaken infrequently [2]. 2.2 The Utility Function The measure of the satisfaction level of an enterprise, and the explicit definition of its goals, are achieved with the utility function U(M;P 1 ; PNp ) where M stands for the available money and P 1 ; PNp represent the parameters which influence the ....

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, March 1989.

.... and architectures: traditional theoretical computer science community [Gal83, Lam78] and in particular computational models and languages communities [Hoa85, Mil89] parallel computation community, distributed algorithms community [Lyn96, Ray88, Tel94, Lam90] distributed artificial intelligence [Dur89, Rum86], operating systems [Ben93, Kis92] and client server research and development community [Dow98] Unfortunately none of the already proposed approaches are applicable to wireless ad hoc networks. For example, parallel computation research is mainly concerned with exploiting concurrency. In addition ....

E.H. Durfee, V.R. Lesser and D.D. Corkill, Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering, Vol.1, pp. 63-83, March 1989.

....the agent could interact with other agents to solve problems. In this respect, an individual agent possesses the power of a group a union set of capabilities of all agents in the group plus additional collaborative behavior. Cooperation is one of the fundamental aspects in DAI #Decker 1987# #Durfee, Lesser, Corkill 1989# #Chaib Draa et al. 1992#. Previous researches on cooperation focus on distribution and collaboration of a given global task #Huhns 1987# #Bond Gasser 1988# #Gasser Huhns 1989# and coordination of individual autonomous agents in a multi agentworld #Demazeau Muller 1990# #Demazeau Muller ....

Durfee, E. H.; Lesser, V. R.; and Corkill, D. D. 1989. Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering 1#1#.

....different disciplines have been proposed for the term multi agent system. As seen from DAI, a multi agent system is a loosely coupled network of problem solver entities that work together Page 4 to find answers to problems that are beyond the individual capabilities or knowledge of each entity [Durfee, Lesser Corkill, 1989]. More recently, the term multi agent system has been given a more general meaning, and it is now used for all types of systems composed of multiple autonomous components showing the following characteristics [Jennings, Sycara Wooldridge, 1998] each agent has incomplete capabilities to ....

Durfee, E.H., Lesser, V.R. and Corkill, D.D. Trends in Cooperative Distributed Problem Solving. In: IEEE Transactions on Knowledge and Data Engineering, March 1989, KDE-1(1), pages 63-83.

.... The clarity of the system actions and representation. System state in the case of failure. To achieve coherence it is necessary to have good coordination to control the interaction between agents within a system to perform the system s goals. To achieve an effective coordination we need [6] : Structure, because without structure agents are not able to interact in predictability ways. Flexibility. Normally agents are a part of a dynamically changing environment which means that flexibility is necessary. Knowledge and reasoning capabilities to use the structure and ....

E. Durfee et al : "Trend in Cooperative Distributed Problem Solving", IEEE Transaction on Knowledge and Data Engineering, Vol. 1 no 1, pp 63-83 March 1989.

....problemsolving skills enabling them to act intelligently on behalf of the entities they represent within the problem domain. A multiagent system can be defined as a loosely coupled network of problem solvers that work together to solve problems that are beyond their individual capabilities [25]. DAI techniques have been successfully applied to a number of telecommunications domains [26, 27] Business services operate in an open, heterogeneous and decentralised management environment in which both control and data are distributed. Agent technology therefore lends itself to this domain. ....

Durfee E H, Lesser V R and Corkill D D: `Trends in co-operative distributed problem solving', IEEE Trans Knowl Data Eng, 11, No 1, pp 63---83 (1989).

....no one agent can (or is willing to) contribute a new state to the problem solving chain. 3.2. Organization as a pool of resources The concept of cooperative problem solving (CDPS) approaches the integration of existing single problem solving experts (intelligent agents) into an overall framework [Durfee, Lesser Corkill 1989]. The aim is to make synergetic use of their individual 10 abilities which, otherwise, could only be used locally. This bottom up perspective of building up a CDPS system is accompanied by a top down perspective on coordinating global processes of problem solving. This approach can ....

....already been a lot of research in the field of individual and joint intentions [Jennings 1992] 3.3. Organization as partitioning of problem spaces 3.3.1. Distributed problem solving Early work within Distributed AI has been concerned with distributed problem solving [e.g. Davis Smith 1983, Durfee, Lesser Corkill 1989, Erman et.al. 1980] The basic approach is shown in fig. 1. Based on this, a diversity of blackboard and contract net architectures have been developed. task decomposition subtask solutions synthesis of subsolutions input: task definition output: overall result Fig. 1: Phases of ....

Durfee, E.; Lesser, V.; Corkill, D.: Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering, Vol. 1, No. 1, 1989, pp. 63-83.

.... of others to try and ensure the community acts in a coherent manner (see [Jen96] and the references within) Cooperative Distributed Problem Solving (CDPS) is a loosely coupled network of problem solvers (nodes) working together to solve problems that are beyond their individual capabilities [DLC89] CDPS targets dicult problems where it is virtually impossible for any one person to conceptualise the whole 3 problem or entire realm of solutions. It often targets problems where even potential solutions are not possible to nd a priori, requiring a solution that must evolve as the situation ....

E.H. Durfee, V.R. Lesser, and D.G. Corkill. Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering, KDE-1(1):63-83, Mar 1989. ftp://ftp.eecs.umich.edu/people/durfee/ieee-tkde89.ps.Z. 11

....reasoning method, and evidence which is used to derive the solution. How to synthesize these multiple solutions from different ESs or agents to produce the final solution for the problem solving is one of critical issues in a cooperative team in the field of distributed expert systems (DESs) [1]. Briefly, the synthesis cases can be classified into belief conflict synthesis cases and non conflict synthesis cases based on the relationships among original evidence sets of solutions in DESs [11] Non conflict synthesis cases include three different cases, which are inclusion, overlap, and ....

....1: The basic frame work of the strategy This integrative strategy works in Probability model [2] and working range of uncertainties is in the area of [0; 1] The structure of the strategy includes four major modules: a) Transformation. If the range of uncertainties of a proposition is not in [0,1], the uncertainties of the proposition are transformed from that range to the range of [0, 1] by using the heterogeneous transformation functions (see [7] For example, if an ES uses the EMYCIN model, the range of [ Gamma1; 1] should be transformed into the range of [0, 1] b) Normalization. ....

[Article contains additional citation context not shown here]

E. Durfee, V. Lesser, and D. Corkill, Trends in Cooperative Distributed Problem Solving, IEEE Transactions on Knowledge and Data Engineering, Vol. KDE-1, No. 1, pp. 63-83.

....architecture (shaded modules represent components particular to agents in a MAS) This short overview article is meant to give the reader a feel for core research issues in the field. The interested reader should also look up other more extensive, though somewhat outdated, reviews of the field [3, 4, 5]. Related reviews on intelligent software agent literature contain more recent references to MAS work [6, 7] In Figure 1 I present an architecture of an intelligent agent situated in a multiagent environment. I have highlighted parts of the architecture which are unique to agents in a multiagent ....

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, March 1989.

....cooperate with the other ESs in any particular problem solving procedure through communication. Each front end processor in HECODES is also equipped with a communication module that manages the communication between itself and other nodes. Cooperation is the central research topic for most DESs [13, 14, 15], and HECODES is no exception. HECODES is used to build DESs to facilitate cooperation among ESs. Four types of cooperation are supported according to the type of relationship involved in the interdependence of ESs. They are: horizontal cooperation; tree cooperation; recursive cooperation; and ....

E. H. Durfee, V. R. Lesser, & D. D. Corkill, Trends in cooperative distributed problem solving, IEEE Transactions on Knowledge and Data Engineering, KDE-1, pp. 63-83, 1989.

....system is a series of parallel agents interacting with each other to solve problems. It is a divide and conquer methodology in which a complex problem is broken up into manageable pieces and solved in parallel. A review of these techniques can be found in Bond and Gasser [3] and Durfee et al. [6]. Distributed systems are more difficult to implement than linear systems, and provide scope for a wider and more subtle variety of bug. But once in place, they provide advantages over conventional control systems both on and off line. Distributing a system achieves 1 two positive effects. ....

E. Durfee, V. Lesser, and D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1:63--83, 1989.

....of (a or b) generation of ( a or b) and (a or b) broker for and broker for a broker for b broker for or generation of a generation of b Figure 5: Message flow in the local caching protocol with four broker agents. For a conceptual characterization of Distributed Problem Solving see [12, 13]. The well known Contract Net Protocol [31] is an opportunist coordination approach using negotiations among distributed agents as its central interaction mechanism for DPS.More recently, a cooperative information gathering approach using a multi agent system based on DPS has been illustrated in ....

Durfee, E. H., Lesser, V. R., Corkill, D. D.: Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering 1:1, 63--83 (Mar. 1989)

....manager agent awards the contract for solving these subtasks to the most appropriate worker agent based on these bids. Contracting involves an exchange of information between agents, an evaluation of the information by each member from its own perspective, and a final agreement by mutual selection [11]. The CNP protocol provides common message formats and a shared communication structure the nodes know about the order of message exchange to generate contracts as well as a specific language which nodes use when exchanging information. The CNP type of negotiation is fairly elementary as the ....

....information accumulated during agent interaction. In multistage negotiation, agents exchange information so that a consensus based on individual constraints is achieved. There is no requirement that each node have a global view of the agents choices and their resource utilization re quirements [11]. Another interesting approach to negotiation is a recent variant of the CNP, called the partial global planning (PGP) 11] In PGP, each agent can represent and reason about the actions and interactions of groups of nodes and how they affect its local activities. In particular, an agent uses its ....

[Article contains additional citation context not shown here]

E. Durfee, V. Lesser, D.D. Corkill "Trends in Cooperative Distributed Problem Solving", IEEE Trans. on Knowledge and Data Engineering, vol.1, no. 1, March 1989.

.... artificial intelligence community has relied on this topdown view and has followed the traditional cognitivist paradigm, that employs explicit representations of goals, beliefs and actions, from un objective observer s viewpoint : a good review of applications to problem solving may be found in (Durfee et al. 1989). This approach has focused on the subproblems of goal identification and communication, conflict resolution, negotiation etc. see several papers in (Bond Gasser 1988) Huhns 1987) Gasser Huhns 1989) Demazeau M ller 1990) and (Demazeau M ller 1991) The same top down cognitivist ....

Durfee, E.H., V.R. Lesser, D.C. Corkill (1989). Trends in cooperative distributed problem solving, IEEE Transactions on Knowledge and Data Engineering, 1(1) :63-83, March 1989.

....place on an agent s choice of actions, the constraints that an agent s commitments to others place on its own choice of actions and the unpredictable evolution of the world caused by other, un modeled agents. Most early work in DAI dealt with groups of agents pursuing common goals (e.g. [90, 38, 21, 91]) Agent interactions were guided by cooperation strategies meant to improve A ROADMAP OF AGENT RESEARCH AND DEVELOPMENT 289 their collective performance. In this light, early work on distributed planning took the approach of complete planning before action. To produce a coherent plan, the ....

V. R. Lesser, E. H. Durfee, and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Transaction on Knowledge and Data Engineering, 1(1):63--83, 1989.

....dispersed and have communication bandwidth. As pointed out by Durfee et al. typical CDPS applications involve problems that consist of a set of interdependent subproblems that arise because of spatial, temporal, and functional distribution of data, knowledge, and processing capabilities [15]. In these applications, a central problem solver is considered as inappropriate because of issues such as limited computation, limited communication, and reliability. CDPS provides techniques to achieve a balance between problem solving and coordination that leads to acceptable overall system ....

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, 1989.

....sorry[cause] Predictor ask[flights] tell[prediction] ask[dd.ad,destination] tell[flight] sorry[cause] Fig. 5. Event flow diagram The second phase consists of analysing the interactions for getting more flexibility (relaxing for example the user requirements) taking advantage of the parallelism [7], duplicating tasks using different methods or resolving detected conflicts. When a cooperation protocol is needed, we should consult the library of cooperation protocols and reuse a protocol definition. If there is no protocol suitable for our needs, it is necessary to define a new one. We can ....

E. H. Durfee, V. R. Lesser, and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1), March 1989.

....function; ffl Agent architecture (degree of heterogeneity, reactive vs. deliberative) ffl System architecture (communication, protocols, human involvement) A useful contribution is that the dimensions are divided into agent and system characteristics. Other overviews of DAI and or MAS include [6, 9, 10]. The taxonomy presented in this article is organized along the most important aspects of agents (as opposed to domains) degree of heterogeneity and degree of communication. Communication is presented as an agent aspect because it is the degree to which the agents communicate (or whether they ....

E. H. Durfee, V. R. Lesser, and D. D. Corkill, "Trends in cooperative distributed problem solving," IEEE Transactions on Knowledge and Data Engineering, vol. 1, pp. 63--83, March 1989.

....an agent manager (so called internal agent in (Rumbaugh, 1995a) of the resource can be suitable. This heuristic has the disadvantage of increasing the dependencies between agents and the overload of interactions (Bond Gasser, 1988) ffl Achieving some of the generic goals of cooperation (Durfee et al. 1989): duplicating tasks with different performing methods. If a task can be achieved by different problem solving methods, a common usage is to draw a goal graph with all the possibilities (Maurer Paulokat, 1994; Mylopoulos et al. 1992) After drawing this diagram, several strategies can be ....

....protocol such as contract net is desired. This can be represented in SDL using the names of the agents or group names in the explicit addressing facility. 8. Determine if a cooperation protocol is needed for each conversation. The reasons for using a cooperation protocol can be, among others (Durfee et al. 1989): ffl Increasing task completion through parallelism. ffl Increasing the set or scope of achievable tasks by sharing resources (information, expertise, physical devices, etc. ffl Increasing the likelihood of completing tasks by undertaking duplicate tasks. ffl Decreasing the interference ....

Durfee, E. H., Lesser, V. R., & Corkill, D. D. (1989). Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1).

....each team member to adhere to the basic joint action convention. In some applications it may be desirable to have more sophisticated joint action conventions, but in every case they must still incorporate these fundamental ideals. 3. 2 Commitments and Conventions: The Cornerstones of Coordination Durfee et al. 1989) identify three major ingredients which must be present for successful coordination: i) there must be structures which enable the agents to interact in predictable 4. This stipulation covers the situation in which an agent which is initially committed to the joint action decides to leave the ....

Durfee, E. H., Lesser, V. R., and Corkill, D. D, (1989) "Trends in Cooperative Distributed Problem Solving" IEEE Trans. on Knowledge and Data Engineering 1 (1) pp 63-83.

.... of the nature of either the individual processes or the particular domain being tackled (Clearwater et al. 1991) This paper does not aim to provide comprehensive coverage of the entire field of DAI such 5 reviews can be found in (Bond and Gasser, 1988; Chaib Draa et al. 1992; Decker, 1987; Durfee et al. 1989; Gasser, 1991 92a; Hern, 1988) Rather the objective is to carry out an in depth analysis on work related to coordinating the problem solving of multiple agents which is one of the central problems of DAI research. At present there are a diverse range of techniques which can and do facilitate ....

....or flat and are responsible for designating the relative authority of the agents and for shaping the types of social interaction which can occur. Organisational structure can also be used as a high level specification of the distribution of problem solving capabilities amongst community members (Durfee et al. 1989). For example when building a community of agents for diagnosing faults in an electricity network (Aarnts et al. 1991; Cockburn et al. 1992) the system designer may specify a functional organisation (agent 1 works on high voltage faults, while agent 2 works at the low voltage level) or a ....

[Article contains additional citation context not shown here]

Durfee, E H, Lesser, V R, and Corkill, D D, 1989. "Trends in Cooperative Distributed Problem Solving", IEEE Trans. on Knowledge and Data Engineering 1 (1) pp 63-83.

....by Durfee, Negotiation is the term used in distributed problem solving research to denote the process by which autonomous nodes coordinate their views of the world and act and interact to achieve their goals. 39] Durfee was involved in early work on coordinating distributed problem solvers [33, 40, 41, 34, 36, 43, 42, 38, 35, 37, 44]. The blackboard model originally arose from the HEARSAY II speech understanding systems which were developed between 1971 and 1976 [93] The HEARSAY II project developed a speech recognition for queries made of a database. HASP, an application of ocean surveillance software for passive sonar, was ....

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, March 1989.

....it would be best to make use of both protocol schemes, or even devise hybrid schemes. Some examples, numerical values for the measure of reuse, and graphical illustrations of reuse potentials are provided. For a conceptual characterization of Distributed Problem Solving see [ Decker et al. 1989; Durfee et al. 1989 ] However, so far DPS has been lacking a real computational model. Our contribution can be seen as step in the direction of providing a computational framework for DPS. More recently, a cooperative information gathering (IG) approach using a multiagent system based on DPS was illustrated in [ ....

E. H. Durfee, V. R. Lesser, and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1, 1, 63--83, March 1989.

....actions that will maximize feedback. Multiagent systems are a particular type of distributed AI system (Bond and Gasser, 1988; Lesser, 1995) in which autonomous intelligent agents inhabit a world with no global control or globally consistent knowledge. In contrast to cooperative problem solvers (Durfee et al. 1989), agents in multiagent systems are not pre disposed to help each other out with all the resources and capabilities that they possess. These agents may still need to coordinate their activities with others to achieve their own local goals. They could benefit from receiving information about what ....

....and out6 of date information. Research in the area of multiagent systems has produced techniques for allowing multiple agents, which share common resources, to coordinate their actions so that individually rational actions do not adversely affect overall system efficiency (Bond and Gasser, 1988; Durfee et al. 1989; Gasser and Huhns, 1989; Huhns, 1987) Coordination of problem solvers, irrespective of whether they are selfish or cooperative, is a key issue to the design of an effective multiagent system. The search for domainindependent coordination mechanisms has yielded some very different, yet effective, ....

Durfee, E. H., Lesser, V. R., and Corkill, D. D. (1989). Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83.

....as cooperative distributed problem solvers, have been shown to be a promising approach. They have been introduced in the domains of manpower management and manufacturing where basics of Distributed Artificial Intelligence, such as contract net protocol or global local planning, have been studied [1]. Following are the more recent of the systems: DAS [2] CORTES [3] ReDs [4] and ADDYMS [5] All of them use a multi agent architecture to address the problem of manufacturing scheduling. However, none of these systems link the task of planning, scheduling and resource allocation into one ....

E. Durfee, V. Lesser and D. Corkill. "Trends in Cooperative Distributed Problem Solving", IEEE Transactions on Knowledge and Data Engineering, Vol I.1, March 1989.

....markets, such as MAGMA [Tsvetovatyy Gini, 1996] employ a global blackboard for posting buying and selling prices. However, such an approach is impractical in realistic applications because it is very difficult to create such a central controller informed of all agents intentions and beliefs. Durfee et al. (1989) point out that such centralised control as in the master slave technique is contrary to the basic assumptions of DAI. 4.3.2 Contracting for Coordination The Contract Net Protocol The most renowned coordination technique for task and resource allocation among agents and determining ....

Durfee, E., Lesser, V. and Corkill, D., Trends in Cooperative Distributed Problem Solving, IEEE Knowledge & Data Engineering, 1, No. 1, pages 63-83, (1989).

....is the motivation behind a significant proportion of current research in the multi agent system arena indeed Gasser (1991) and Hewitt (1991) identify it as perhaps the key problem in this area. To date, a diverse range of solutions have been adopted ranging from organisational structures (Durfee et al. 1989; Gasser et al. 1989; Shoham and Tennenholtz, 1992; Werner, 1989) to meta level information exchange (Durfee, 1988; Gasser, 1992a) to multi agent planning (Cammarata et al. 1983; Corkill, 1979; Georgeff, 1983; Rosenschein and Genesereth, 1985) Whilst all of these mechanisms offer some insight ....

Durfee, E. H., Lesser, V. R., and Corkill, D. D, (1989) "Trends in Cooperative Distributed Problem Solving" IEEE Trans. on Knowledge and Data Engineering 1 (1) pp 63-83.

....one representative trial. 14 Figure 7: Similarity of the rectangle and pick and place trajectories. Tech. Report CMU RI TR 93 11 iv 1 Introduction Distributed problem solving (DPS) has had a long history in computer science [1][2], but, historically, little of this work was applied to hard real time systems. Behavior based frameworks, a relative new comer to the DPS field, have made strong contributions to the real time world of robotic control as evidenced by the subsumption architecture [3] and others [4] 5] The success ....

Durfee, E.H., V.R. Lesser and D.D. Corkill, "Trends in Cooperative Distributed Problem Solving," IEEE Transactions on Knowledge and Data Engineering, v.1, n.1, March 1989, pp. 63-83.

....their reasoning and planning. For coordination to be successful when there are these forms of uncertainties, there must be structures that allow agents to interact predictably, and also flexibility for dynamic environment and imprecise viewpoints, in addition to the local reasoning capability [5]. For this purpose, we propose a domain independent, flexible negotiation framework for the agents to negotiate their commitments. Our work differs from the conventions and social conventions [8, 9] in that our negotiation framework is domainindependent, and allows the agent to integrate the ....

E. H. Durfee, V. R. Lesser, and D. D. Corkill. Trends in cooperative distributed problem solving. In IEEE Trans. Knowledge Data Eng. 1(1):63-83, 1989.

....their reasoning and planning. For coordination to be successful when there are these forms of uncertainties, there must be structures that allow agents to interact predictably, and also flexibility for dynamic environment and imprecise viewpoints, in addition to the local reasoning capability [Durfee, Lesser, and Corkill 1989]. For this purpose, we propose a domain independent, flexible negotiation framework for the agents to negotiate their commitments. Our work differs from the conventions and social conventions [Jennings 1993, Jennings 1996] in that our negotiation framework is domain independent, and allows the ....

E. H. Durfee, V. R. Lesser, and D. D. Corkill, "Trends in cooperative distributed problem solving". IEEE Trans. Knowledge. Data Eng. 1(1):63-83, 1989.

....among the processes controlling the separate resources. Distributed resource allocation, therefore, involves the cooperative solution of resource allocation problems among a network of decision makers, and falls into the subclass of DAI known as Cooperative Distributed Problem Solving (CDPS) (Durfee, Lesser, and Corkill 1989). Individual agents in a Distributed Meeting Scheduling (DMS) system have only partial knowledge of system wide goals because they do not know all the meetings that are currently being scheduled or have already been scheduled by other agents. Hence they must exchange relevant information to ....

Durfee, Edmund H., Victor R. Lesser, and Daniel D. Corkill. (1989). Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83.

No context found.

E. Durfee, V. Lesser, and D. Corkill, "Trends in cooperative distributed problem solving," IEEE Transactions on Knowledge and Data Engineering, vol. KDE-1, no. 1, pp. 63--83, March 1989.

No context found.

E. Durfee, V. Lesser, & D. Corkill. Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering. 1995.

No context found.

E. Durfee, V. Lesser, & D. Corkill. Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering. 1995.

No context found.

E. H. Durfee, V. R. Lesser, and D. D. Corkill. Trends in cooperative distributed problem solving. IEEE Transaction on Knowledge and Data Engineering, 1:63 -- 83, 1989.

No context found.

E.H. Durfee, V.R. Lesser, and D.D. Corkill. Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, March 1989.

No context found.

Durfee, E. H., Lesser, V. R. and Corkill, D. D. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering,1, 63-83 (1989).

No context found.

Durfee, E.; Lesser, V.; Corkill, D.: Trends in Cooperative Distributed Problem Solving. IEEE Transactions on Knowledge and Data Engineering, Vol. 1, No. 1, 1989, pp. 63-83. -

No context found.

Durfee E. H., V. R. Lesser and D. C. Corkill (1989). Trends in cooperative distributed problem solving, IEEE Transactions on Knowledge and Data Engineering, 1(1):63-83, March 1989.

No context found.

Edmund H. Durfee, Victor R. Lesser, and Daniel D. Corkill. Trends in cooperative distributed problem solving. IEEE Transactions on Knowledge and Data Engineering, 1(1):63--83, March 1989.

First 50 documents  Next 50

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