Last updated September 21, 1996

How to use Alta Vista

Go to Alta Vista

Quick Search Form:
	Input
	Sort On
Abstract: The Alta Vista search engine developed by Digital Corp. is one of the top three general subject search engines on the Internet. It handles over 4 million search requests a day. Although the WWW index, alone, is over 30 gigabytes in size, a search generally takes less than a few seconds. The indexes which form the basis of the system are generated by a fast web crawler that scans roughly 2.5 million WWW pages a day. The system also indexes Usenet news databases daily. The advanced query interface allows both Boolean and proximity operators. It allows the search output to be sorted by keyword. Compact and detailed output formats are available as well as a simple hit count. On the basis of our tests Alta Vista finds a directly relevant hit within the first ten sites listed when results are sorted. Because it is web crawler-based there is no discernable slant in the overall coverage with regard to either the type of information indexed or the types of sites indexed.

How to use Excite!

Go to Excite!

Quick Search Form:
	Input
Abstract: The Excite! search engine developed by Excite Inc. is another of the general WWW subject search engines. The Excite! search indexes are regenerated weekly by a web crawler thereby minimizing dead links. Search results are ranked by relative confidence score and displayed with brief summaries. The system offers a 'Query by Example' button in the search output that performs a secondary search on the terms in the selected item. Users can search the WWW, Usenet or Excite's own NetReviews using either the simple search interface or the advanced search interface. The advanced search interface permits the use of standard Boolean operators. In addition to the general search engine Excite offers a NetDirectory service which can be browsed by category or searched. Selected sites are reviewed by Excite! editorial staff. Review summaries are prepared manually. These reviews form the basis of Excite's NetReviews service. Users may suggest or contribute a site for inclusion in NetReviews or submit their URLs for analysis by the Excite! web crawler.

---

Infoseek

---

How to use Infoseek

Go to Infoseek

Quick Search Form:
	Input
Abstract: Although not having the biggest database among the primary search engines on the Web, the Infoseek search engine, produced by Infoseek Corporation, claims to be the most accurate and to have the smartest search tools. The full text of each page in the Infoseek database is indexed, searches are case-sensitive, and proximity rankings are used to improve the relevancy of its findings. Also, the search results which are displayed are very complete, offering the Web page title, its URL, a relevancy score, the file size, and a computer-generated summary. Unfortunately, since Infoseek indexes only about one million web pages, it is far from comprehensive. Users will need to extend their search using other search engines if in-depth information is required.

Authors' Note: The Inktomi search engine has been
replaced by the HotBot search engine.

See the 1997 review for details.

R. D. Ralph 4/18/97

Quick Search Form:
	Input
Abstract: Inktomi is a research prototype site created by the University of California at Berkeley. It will soon be replaced by a "fully featured commercial search service." It currently lacks much of the graphics-intensive mapping that slows down other similar engines. While its database is smaller than other systems, such as Lycos and OpenText, Inktomi covers its corner of the information universe, particularly academics, very well. The search engine does not allow Boolean searching. Inktomi searches each term entered in a multi-term query placing an implicit Boolean OR between all terms entered. It matches roots of search terms, so that the search term "walking" would generate hits under "walk" and "walker" as well. Results are ranked by a relevancy score, and may be returned in normal or terse text, or in a tediously slow and pointless graphics mode.

How to use Lycos

Go to Lycos

Quick Search Form:
	Input
Match: Loose Fair Good Close Strong
Abstract: Calling itself the "Catalog of the Internet," Lycos' database containing over 37 million URLs gives some credence to this claim. Begun at Carnegie Mellon University but now independent, Lycos offers several options to the searcher, which enable Boolean searching, establish a relevancy standard for the search, and allow displaying the results in a variety of formats. Returned search results are ranked by a relevancy score and a computed adjacency index. Attached to Lycos are two other services: A2Z, which purports to abstract every site in the Lycos directory, and Point, which seems to provide an identical service. Of the two, A2Z is easier to use because it provides a brief abstract of each site with a direct link.

How to use OpenText

Go to OpenText

Quick Search Form:
	Input
Abstract: The actual search engine behind Yahoo!, the OpenText Index offers the best search interface among the search engines currently available. The form allows for Boolean and proximity searching, as well as searching the entire site, rather than keywords alone. All words in a site document are indexed by the OpenText web crawler. The database OpenText uses has almost 10 billion words indexed on approximately the same number of sites Lycos indexes (37 million). Results of searches are ranked by relevancy. OpenText's most powerful feature is the "See matches on the page" feature. It generates a KWIC index on-the-fly which allows the searcher to see the context in which the search terms appear in the returned site document. This is very helpful in screening the result for false drops.

How to use WebCrawler

Go to WebCrawler

Quick Search Form:
	Input
Abstract: At first glance, WebCrawler appears to be the lean, quick-response database which was initially a research project developed at the University of Washington in April 1994. But since its acquistion by America Online, Inc. in April of 1995, and its integration with Global Network Navigator's "Personal Library" search software, WebCrawler is decidedly a larger, more powerful, and more intuitive search engine. In addition to natural language searching, WebCrawler supports a full range of Boolean operators including adjacency, phrase only, and combined NOT/OR searching. Other new, special features include the option of reviewing search results in a short or detailed format, no-forms search support, and full integration with GNN Select's Web review pages. URL submissions are now accepted for next-day inclusion into the WebCrawler index. Handling over 3 million queries a day, WebCrawler has regained its status as a powerful, comprehensive search engine quite capable of providing relevant results for the most complex, challenging searches.

How to use Yahoo!

Go to Yahoo!

Quick Search Form:
	Input
Abstract: Yahoo!, begun at Stanford University but now an independent entity, is an hierarchical, subject-oriented guide to "tens of thousands" of sites on the 'Net. The search engine behind Yahoo! is OpenText but not all advanced features are engaged. Searches can be modified to query only the Yahoo! directory or the whole Internet, including Usenet newsgroups and e-mail addresses. Only Boolean operators AND and OR can be used and there is no option to limit or enhance search results. Yahoo!'s directory is one of the first attempts at categorizing information found on the Internet and is still considered a superb "starting point." However, if a complex search or specific, detailed information is sought, it is better to use an advanced search engine such as OpenText or Alta Vista.

These Pages Represent a Class Project of
LIS646 - Abstracting and Indexing - Spring '96
Dr. Randy D. Ralph, Instructor
Department of Library and Information Studies
School of Education
University of North Carolina at Greensboro

Contact the Participants
Randy D. Ralph	John W. Felts	Ben J. Lea