Schwartz, M. F.; The Networked Resource Discovery Project; In Proceedings of the IFIP XI Worm Congress; San Francisco, California, 1989; pp 827- 832.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....to read time is called retrieval, and is addressed in our research by the HELGON system [5] Adding personal retrieval semantics at storage time helps make retrieval easier. When users have a question but have not previously stored any relevant information, they engage in a process of discovery [13], in which a new message is sent as a request for augmentation of one s own environment. Because the process of discovery involves interaction with many agents (the other users of Usenet News) it is highly unlikely that any two identical queries will result in the same set of answers. This makes ....

Schwartz, M. F.; The Networked Resource Discovery Project; In Proceedings of the IFIP XI Worm Congress; San Francisco, California, 1989; pp 827- 832.

....work Previous work can be broken down into the following broad categories: distributed naming systems, network navigation systems and network based information retrieval systems. Distributed naming systems such as X. 500 [CCI88] Profile [Pet88] and the Networked Resource Discovery Project [Sch89] provide attribute based access to a wide variety of objects; however, they do not support the hierarchical relationship between servers and data which our system achieves 3 through its use of content labels. Also, the systems are not integrated into a file system, nor do they provide automatic ....

Michael F. Schwartz. The Networked Resource Discovery Project. In Proceedings of the IFIP XI World Congress, pages 827--832. IFIP, August 1989.

....using the partitioning criteria of a relation. We prefer a more general approach that uses information about data distribution patterns to improve performance and does not require owners of data to constrain the values of attributes in other organizations. The Networked Resource Discovery Project [13, 14] provides an architecture for locating a few instances of a type of resource when the resource type is prevalent in the network. It multicasts queries to a probabilisticly chosen subset of the available data repositories. Successive queries do not return the same answer, and queries may fail even ....

M. F. Schwartz, "The Networked Resource Discovery Project," IFIP XI World Congress, San Francisco, pp. 827-832 (August 1989).

....through browsing or the use of simple, keyword based search. A number of schemes focussing on improved resource location within an internet environment have been proposed. One such scheme is described below. 2.2.4. 2 The Networked Resource Discovery Project The Networked Resource Discovery Project [Sch89, Sch90b] has investigated means by which users can discover the existence of a variety of resources in an internet environment. The focus is on supporting attribute based Yellow Pages searches. The information repositories in the system register a high level description of the information which ....

Michael F. Schwartz. The networked resource discovery project. In IFIP XI World Congress (San Francisco), pages 827--832, Aug 1989.

....network. It starts its search by looking for the organization name in a database built from network news articles. It then uses the Internet Domain Naming System to find hosts to look for and finally the finger protocol to search for the individual. Other approaches include user discovery agents [Schwartz 89] which accept queries from users and use the information provided in the query to find objects in which the user has an interest. The information needed to direct a query to the appropriate agent evolves over time. A query is directed to the nearest agent, and agents learn how to direct queries ....

M. F. Schwartz. The networked resource discovery project. In Proceedings of the IFIP XI World Congress, pages 827--832, August 1989. 125

....The key difference is that the links do not necessarily form a hierarchy. Alternative approaches are being examined by the Resource Discovery Project at the University of Colorado. These approaches use information already available over the network. In one approach, resource discovery agents [28] collect and share information with other agents scattered across the system. A user wishing to find a resource asks one of these agents, and the agents route queries among themselves, exploiting the semantics of the query to limit the activity that must take place. If the resource is found, a ....

M. F. Schwartz. The networked resource discovery project. In Proceedings of the IFIP XI World Congress, pages 827--832, August 1989. San Francisco.

.... processes out to scour the network and return with useful tidbits of information [10] One name service snoops in news headers in an effort to locate people [20] Other name services look around and return some answers to a query, but not all answers and not necessarily the same ones as last time [19, 20]. We could describe the semantics of this type of query processing as luck semantics. Even if we must depart from the traditional model of consistency, we would like to improve on luck by supplying results that can be interpreted and used without frustration. Traditional consistency has three ....

M. F. Schwartz, "The Networked Resource Discovery Project," IFIP XI World Congress, San Francisco, pp. 827-832 (August 1989).

.... the distinction refers to the domain of the objects named by descriptive services (services or people) Schwartz argues that the distinction between yellow and white pages services is that yellow page queries can be answered with information about a few of many available matching objects [92]. Allowing users to supply limits on the size of a result accomplishes this query for all types of descriptive name services. Efforts to optimize size limits could be made for those situations (network services) where a size limit is most likely to occur. 2.2. Scale The scale of the global ....

Schwartz, M. F., "The Networked Resource Discovery Project," IFIP XI World Congress, San Francisco, pp. 827-832 (August 1989).

....as asking the appropriate person) to be included as part of the file system. This information provides a matrix through which users can navigate to find the desired information. Services might even spring up to help users navigate through this matrix. An example of such a service is described in [14] and makes use of resource discovery agents. These agents accept queries from users and use the information provided by the user to find objects in which the user is interested. The multitude of links in a system based on the virtual system model can provide the information needed to direct such ....

M. F. Schwartz. The networked resource discovery project. In Proceedings of the IFIP XI World Congress, pages 827--832, August 1989. San Francisco.

....previous system combines a coherent architecture for associative access to distributed, heterogeneous information servers with query specific feedback to help the user organize the search space. Distributed naming systems such as X. 500 [4] Profile [16] the Networked Resource Discovery Project [17], and Nomenclator [14] provide attribute based access to a wide variety of objects; however, they do not support the hierarchical relationship between servers and data that our system achieves through its use of content labels. Network navigation systems such as the Gopher system [1] and the ....

M. F. Schwartz. The Networked Resource Discovery Project. In Proceedings of the IFIP XI World Congress, pages 827--832. IFIP, Aug. 1989.

....name for the hierarchy can only be added at the point of attachment. Below that level, all names remain the same as in the primary hierarchy. There are other approaches that have been proposed for identifying objects of potential interest. Schwartz proposes the use of resource discovery agents[10] which accept queries from users and use the information provided by the user to find objects in which the user is interested. In Schwartz s design, the information needed to direct a query to the appropriate agent evolves over time. A query is directed to the nearest agent, and agents learn how ....

M. F. Schwartz. The networked resource discovery project. In Proceedings of the IFIP XI World Congress, pages 827--832, August 1989. San Francisco.

....The Networked Resource Discovery Project is investigating means by which users can discover the existence of a variety of resources in an internet 2 environment, including retail products, network services, and people in various capacities. The project involves several goals and techniques [Schwartz 1989]. Our main concerns are flexible organization, scalability, and lack of global administrative authority. These constraints are difficult to satisfy simultaneously. Traditional directory services (such as the CCITT X.500 standard [CCITT 1988] rely on hierarchical organization to achieve good ....

M. F. Schwartz. The Networked Resource Discovery Project. Proceedings of the IFIP XI World Congress, pp. 827-832, San Francisco, California, August 1989.

....environments, spanning national or international networks. This general problem has taken a number of specific forms, including Internet white pages [Schwartz Tsirigotis 1991] support for distributed collaboration [Schwartz Wood 1990] support for probabilistic yellow pages management [Schwartz 1989, Schwartz 1990] and support for mapping and discovering public resources reachable via the Internet [Schwartz et al. 1991] A common theme of these projects is the ability to accommodate heterogeneity of upper level network protocols, information formats, and organizational structures. While ....

M. F. Schwartz. The Networked Resource Discovery Project. Proc. IFIP XI World Congress, pp. 827-832, San Francisco, CA, Aug. 1989.

.... included research into providing Internet white pages [Schwartz Tsirigotis 1991a] characterizing the organizational structure of distributed collaboration via electronic mail [Schwartz Wood 1990] devising probabilistic algorithms for supporting attribute based ( yellow pages ) searches [Schwartz 1989, Schwartz 1990] supporting resource discovery among anonymous FTP sites [Schwartz et al. 1991a] and supporting means of discovering and visualizing characteristics of large internets [Schwartz et al. 1991b] The reader interested in more details about the various projects and prototypes is ....

M. F. Schwartz. The Networked Resource Discovery Project. Proc. IFIP XI World Congress, pp. 827-832, San Francisco, CA, Aug. 1989. -

....dissemination, followed by the cache management protocols that evolve specialization subgraphs. 3.1. Sparse Diffusion Multicast Primitive To support probabilistic information dissemination and searches, we have proposed a primitive called a Sparse Diffusion Multicast (SDM) defined as follows [Schwartz 1989]. Given a network containing N agents, a message is sent using an SDM to n N agents, selected at random. The parameter n determines the SDM 4 density. In addition to SDM density, the SDM primitive is also parameterized by scope. For example, an SDM could be initiated within a city, or ....

M. F. Schwartz. The Networked Resource Discovery Project. Proc. IFIP XI World Congress, pp. 827-832, San Francisco, CA, Aug. 1989.

No context found.

Michael F. Schwartz, "The Networked Resource Discovery Project," Proc. IFIP XI World Congress, pp. 827-832 (Aug. 1989).

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