Udi Manber, Mike Smith, and Burra Gopal. WebGlimpse -- combining browsing and searching. In Proceedings of 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....Then, the user choices one of the proposed services and the corresponding application is run (namely, the corresponding agent is activated) Another example of information management applications that multiagent systems address is the integration of browsing and searching. In fact, WebGlimpse [92] is a software agent that improves the searching of documents that are related to those a user is currently browsing. The system has been developed as a single agent, because the Web is, at the moment, too slow and the bandwidth is too narrow for supporting a large number of mobile agents. Another ....

....on some other operations. Among personal agents, we recall those devoted to provide a graceful interface to the user [51] those devoted to rate Web pages or news articles in order to present only the most pertinent ones to the user [86, 87, 90, 105, 122] those devoted to search information [23, 72, 92], and those devoted to schedule the user s agenda [21, 27, 70, 85] Medical applications. Medical applications include patient monitoring and health care. Merchant et al. 97] describe an application that, exploiting agents that move around a network, performs biomedical simulations. ....

U. Manber, M. Smith, and B. Gopal. Webglimpse - combining browsing and searching. In Proceedings of the Usenix Technical Conference, January 6-10 1997.

....poor in terms of completeness. Linguistic regularities and behavior patterns within the WWW seem to be the fruit of a simple language evolution rule : maximum communication with minimal effort [Ami97] In hypertext systems, the neighborhood of a node is often considered as being its context [Man97] On the WWW, there is a high diversity of link types. One can distinguish not only structural links which establish a contextual relationship between pages (e.g. People, Research, Academic, International Relations, outward links from an university homepage) but also navigational links (e.g. ....

U. Manber. Webglimpse - combining browsing and searching. In In Usenix Technical Conference, 1997.

....a graphical visualization tool on top of AIDE. The AIDE version repository retrieves and stores only pages that have explicitly been requested by users. As such, the size of the repository is typically much smaller than the sizes targeted by WebBase. Similarly, GlimpseHTTP (now called WebGlimpse) [13] provides text indexing and neighborhood based search facilities on existing repositories of web pages. Here again, the emphasis is more on the actual indexing facility and much less on the construction and maintenance of the repository. The Internet Archive [11] project aims to build a digital ....

Udi Manber, Mike Smith, and Burra Gopal, WebGlimpse - Combining browsing and searching, Proc. of the

....a graphical visualization tool on top of AIDE. The AIDE version repository retrieves and stores only pages that have explicitly been requested by users. As such, the size of the repository is typically much smaller than the sizes targeted by WebBase. Similarly, GlimpseHTTP (now called WebGlimpse) [MSG97] provides text indexing and neighborhood based search facilities on existing repositories of web pages. Here again, the emphasis is more on the actual indexing facility and much less on the construction and maintenance of the repository. The Internet Archive [IArch] project aims to build a ....

Udi Manber, Mike Smith, and Burra Gopal, WebGlimpse - Combining Browsing and Searching , Proc. of the 1997 USENIX Technical Conference, Jan 6-10, 1997.

.... for Dynamic Reranking We integrated the dynamic reranking feature described above, in an experimental version of Fetuccino [Ben Shaul et al. 1999] Fetuccino is a search parasite that validates and augments existing Web search services results (by local directed crawling a la WebGlimpse [Manber et al. 1997]) and finally visualizes them in a Java applet. Its standard version (without the rerank feature) is available as a free service on the IBM Corporate Java page 6 , as well as a free stand alone application on the IBM Alphaworks site 7 . The integration was done by simply modifying the ....

Manber, U., Smith, M., and Gopal, B. (1997). Webglimpse: Combining browsing and searching. In Proceedings of the Usenix Technical Conference, Los Angeles, CA.

....sites do not differentiate clients, however many Web sites have their own search engines for internal documents. We assume for such Web sites, clients for these Web sites can be differentiated as registered (class A with weight 2) and unregistered users (class B with weight 1) We use WebGlimpse [27] as keyword searching software. To examine the performance of DDSD, we need a trace and test documents. We have collected about 160,000 HTML documents from a few Web sites and replicate them on all the 8 nodes. For the trace, we use a query log collected from the search engine site www.google.com ....

Udi Manber, Mike Smith, and Burra Gopal, "Webglimpse -- combining browsing and searching," Proceedings of the Usenix Technical Conference, Jan. 1997.

.... Search Parasite for Dynamic Reranking We integrated the dynamic reranking feature described above, in an experimental version of Fetuccino [BSHJ 99] Fetuccino is a search parasite that validates and augments existing Web search services results (by local directed crawling a la WebGlimpse [MSG97] and finally visualizes them in a Java applet. Its standard version (without the rerank feature) is available as a free service on the IBM Corporate Java page 6 , as well as a free stand alone application on the IBM Alphaworks site 7 . The integration was done by simply modifying the ....

U. Manber, M. Smith, and B. Gopal. Webglimpse: Combining browsing and searching. In Proceedings of the Usenix Technical Conference, Los Angeles, CA, 1997.

....sites do not differentiate clients, however many Web sites have their own search engines for internal documents. We assume for such Web sites, clients for these Web sites can be differentiated as registered (class A with weight 2) and unregistered users (class B with weight 1) We use WebGlimpse [26] as keyword searching software. To examine the performance of DDSD, we need a trace and test documents. We have collected about 160,000 HTML documents from a few Web sites and replicate them on all the 8 nodes. For the trace, we use a query log collected from the search engine site www.google.com ....

Udi Manber, Mike Smith, and Burra Gopal, "Webglimpse -- combining browsing and searching," Proceedings of the Usenix Technical Conference, Jan. 1997.

....a graphical visualization tool on top of AIDE. The AIDE version repository retrieves and stores only pages that have explicitly been requested by users. As such, the size of the repository is typically much smaller than the sizes targeted by WebBase. Similarly, GlimpseHTTP (now called WebGlimpse) [MSG97] provides text indexing and neighborhood based search facilities on existing repositories of web pages. Here again, the emphasis is more on the actual indexing facility and much less on the construction and maintenance of the repository. The Internet Archive [IArch] project aims to build a ....

Udi Manber, Mike Smith, and Burra Gopal, WebGlimpse - Combining Browsing and Searching , Proc. of the 1997 USENIX Technical Conference, Jan 6-10, 1997.

....several variants of the crawler indexer architecture. Between them we can mention Harvest [14] which uses a distributed architecture to gather and distribute data, being more efficient. However, the main drawback is that needs the coordination of several WWW servers. Another variant is WebGlimpse [28] that attaches a small search box to the bottom of every HTML page, and allows the search to cover the neighborhood of that page or the whole site, without having to stop browsing. We also have to mention search engines that specializes in specific topics. For example the Search Broker [27] or the ....

U. Manber, M. Smith, and B. Gopal. Webglimpse: combining browsing and searching. In Proc. of USENIX Technical Conference, 1997.

....Intranet Web Server The solution to the apparently trivial case of a single organizational web server uncovers some interesting alternatives. If an organization has a single web server, any number of local search products may be chosen to full text index the document corpus. For example, Glimpse (Manber et al. 1997), WAIS (Kahle, 1995) or Verity would all be perfectly good choices. Implementation Approach: Pre Coordinated Searching The situation grows more interesting if a pre coordinated search is desired. AWeb front end HTML form can be built to submit a document to a staging area along with metadata ....

Manber, U., Smith, M., and Gopal, B. (1997). WebGlimpse: Combining browsing and searching. In Usenix Technical Conference. Usenix.

....2: The outline view of the current implementation of Cha Cha search on the query earthquake . Many web sites show site maps which su#er from the same limitation. Users can browse or navigate the structure of the site, but cannot search within or across that structure. The WebGlimpse system [22] provides crawling and indexing of an intranet (both local and external links) It also allows system administrators to define link neighborhoods in one of two ways: all pages within k links of a given page, or all pages within a file system subdirectory. We explored this approach in Cha Cha ....

Udi Manber, Mike Smith, and Burra Gopal. WebGlimpse -- combining browsing and searching. In Proceedings of 1997 Usenix Technical Conference, 1997.

....WebSQL for application development by describing in some detail two applications, Web site maintenance and specialized index construction, and sketching several others. What sort of applications do we have in mind Consider for example the WebGlimpse system developed at the University of Arizona [MSG97]. WebGlimpse combines searching and browsing by allowing a user to request at any time a full text search, not on the whole Web through some large index server, but on the neighborhood of a given page that has been reached by browsing. For example, if I am looking for Canadian price lists for ....

Udi Manber, Mike Smith, and Burra Gopal. WebGlimpse - Combining browsing and searching. In Proceedings of the 1997 Usenix Technical Conference, 1997. ftp://ftp.cs.arizona.edu/people/udi/webglimpse.ps.Z.

....Intranet Web Server The solution to the apparently trivial case of a single organizational web server uncovers some interesting alternatives. If an organization has a single web server, any number of local search products may be chosen to full text index the document corpus. For example, Glimpse (Manber et al. 1997), WAIS (Kahle, 1995) or Verity would all be perfectly good choices. 2.1.1 Implementation Approach: Pre Coordinated Searching The situation grows more interesting if a pre coordinated search is desired. A Web front end HTML form can be built to submit a document to a staging area along with ....

Manber, U., Smith, M., and Gopal, B. (1997). Webglimpse: Combining browsing and searching. In Usenix Technical Conference. Usenix.

....have a query search facility for file systems. Glimpse is a tool that provides an inverted index facility for file systems with a boolean search capacity [23] WebGlimpse meanwhile is a project to combine browsing and searching over the Web by defining neighborhoods of given documents for browsing [22]. We focus on boolean queries in this paper. We should note, however, that many Web search engines do not provide a boolean query interface as their standard facility. Instead, they process keyword queries by other informationretrieval (IR) techniques. For instance, AltaVista s standard interface ....

U. Manber, M. Smith, and B. Gopal. WebGlimpse--- combining browsing and searching. In Proceedings of the Usenix Technical Conference, Los Angeles, California, Jan. 1997.

....back to the user. This simple minded approach turns out to be extremely powerful. The proliferation of search software, often for free, made it easy to provide search capabilities on many web sites. We are proud to be partially responsible for that with our Glimpse [11] GlimpseHTTP, WebGlimpse [10], and Harvest [2] systems. Within the last year thousands of search servers have been added. Most of them deal with very limited specific information (e.g. they search the content of one site) but many provide professional content in areas of general interest. The trend to connect existing ....

Manber U., M. Smith, and B. Gopal, "WebGlimpse -- Combining Browsing and Searching," Usenix 1997 Annual Technical Conference, Anaheim, CA (January 1997), pp. 195-206.

....this script so that we can control the running time of the script in order to generate different loads by CPU busy spinning. As a result, these CGI requests are CPU intensive. For the KSU library searching requests, we created a database with approximately 10000 items using the WebGlimpse software [27] and replaced the CGI library requests with WebGlimpse commands. These requests have mixed CPU and I O demands, but on average 90 of service time is spent searching index information in memory. For the ADL trace, we replicated a small ADL catalog database in our local cluster and ran an ADL ....

U. Manber, M. Smith, and B. Gopal. Webglimpse -- combining browsing and searching. Proceedings of the Usenix Technical Conference, January 1997.

....addition, each query page contains hypertext links to all the files and subdirectories of the corresponding directory to allow users to browse them. GlimpseHTTP uses Glimpse since it supports very flexible and efficient ways to limit searches to only parts of the indexed information. WebGlimpse [30] is an extension of GlimpseHTTP. It does not assume that the the information base is a UNIX hierarchical file structure. Instead, it interprets the files in the information base as hypertext documents. For each document, WebGlimpse builds a neighbourhood file that contains the names of all the ....

U. Manber, M. Smith, and B. Gopal. WebGlimpse -- Combining Browsing and Searching. To appear in: Usenix Annual Technical Conference. In Jan 6-10, Anaheim, California, 1997.

....this script so that we can control the running time of the script in order to generate different loads by CPU busy spinning. As a result, these CGI requests are CPU intensive. For the KSU library searching requests, we created a database with approximately 10000 items using the WebGlimpse software [28] and replaced the CGI library requests with WebGlimpse commands. These requests have mixed CPU and I O demands, but on average 90 of service time is spent searching index information in memory. For the ADL trace, we replicated a small ADL catalog database in our local cluster and ran an ADL ....

U. Manber, M. Smith, and B. Gopal. Webglimpse -- combining browsing and searching. Proceedings of the Usenix Technical Conference, January 1997.

No context found.

Udi Manber, Mike Smith, and Burra Gopal. WebGlimpse -- combining browsing and searching. In Proceedings of 1997.

No context found.

U. Manber, M. Smith, and B. Gopal. WebGlimpse: combining browsing and searching. USENIX Annual Technical Conference. USENIX Association, 1997.

No context found.

U. Manber, M. Smith, and B. Gopal. WebGlimpse: combining browsing and searching. USENIX Annual Technical Conference. USENIX Association, 1997.

No context found.

U. Manber, M. Smith, and B. Gopal. WebGlimpse: combining browsing and searching. USENIX Annual Technical Conference. USENIX Association, 1997.

No context found.

Udi Manber, Michael Smith, Burra Gopal, "WebGlimpse -- Combining Browsing and Searching", Proc. of USENIX Annual Technical Conf., Anaheim, (January 1997.

No context found.

Udi Manber, Michael Smith, Burra Gopal, "WebGlimpse -- Combining Browsing and Searching", Proc. of USENIX Annual Technical Conf., Anaheim, (January 1997.

No context found.

U. Manber, M. Smith, and B. Gopal, "Webglimpse: Combining Browsing and Searching," Proc. 16th ACM Symp. on Principles of Database Systems, ACM Press, New York, Jan. 1997, pp. 195-206.

No context found.

Manbet, U., Smith, M., and Gopal, B.: WebGlimpse: Combining Browsing and Searching. In Proceedings of the USENIX

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