Using the Web to find Information

Keys to Searching


You've got questions? The Internet has answers. But it only has some of the answers.There is a massive amount of information that is not, and perhaps never will be on the Internet. So if you don't find your answer on the Internet, check your library (Tennant, 1996).

Approaches to Web searching.

Search Engine Lists

These are merely lists of search engines collected in one location. You can access the engines right from the list.

Test out various engines and find one or two that work best for your search needs. Search engines vary in breadth and coverage, that is, the number of Web sites they record. They also vary in depth. One may look only through titles while another may examine opening paragraphs or even full text. In addition, available search techniques differ from one engine to the next. Consequently, you may fine that one engine suits your needs better than another.

c|net (SEARCH.COM)
This site lists 250 Web searchers, all categorized. You may search the searchers to find one appropriate to a particular inquiry. Welcome to search.com describes how to use this search index. You may want to start with this page and then return to the homepage of the site.
The All-in-One Search Page
A Web page which lists nearly 200 search tools.
Useful Internet Resources
Tools for Searching the Internet -- Looking for contact information of a colleague? Mutual fund performance over the last five years? All that and more is increasingly just a few mouse-clicks away with the latest Internet search services. Contains the following search categories.
  • Subject directories
  • Web databases
  • Search engine lists
  • Metaserach Tools
  • The comprehensive search tool
  • Job banks
  • General reference tools
  • Government Information
  • Web sites for higher education
  • Web Sites for mechanical engineering
  • Sites for chemistry
  • Sites for mathematics
  • Sites for biology
  • Sites for nursing and health sciences
  • Sites for business and economics
  • Sites for financial aid

 

The Configurable Unified Search Index (CUSI)
A user-friendly list. It has a single-search form. You enter your search request and CUSI "translates" it into a format acceptable to the search engine you have selected. Thus, you need to become familiar with only one search format and enter a particular search only one time.
Internet Sleuth
A rather amazing list, focusing not on search engines, but on searchable, stand-alone databases. It contains links to more than 1,000 databases. As with c|net, the databases are categorized. You may choose to search within Sleuth to find a database appropriate to a particular topic.
Open Text Web Index
A Web database that provides serveral different search interfaces, depending upon your need. For quick and basic searches there is the "Simple Search." For those who know how to use Boolean operators and would like the additional search syntax control they provide there is "Power Search." And finall, there is "Weighted Search," where the user can assign numberic values that indicate the relative importance of the search words.
Alta Vista
The latest Web database to appear on the Internet also appears to be one of the best. Tennant (1996) reports that the Web robot that builds this database (called"Scooter," incidentally) even located and indexed two documents from his desktop Macintosh. The database claims to index over 16 million Web pages containing over 8 billion words. IT offers a search of this massive web database, or searching the full-text of over 13,000 Usenet news groups. It offers both a simple search and an advanced search option. In the advanced search option you can use Boolean operators and also select additiona words to use ot rank results.

An example of an Advanced Alta Vista Search is provided for you to try.

Metasearch Tools

Erik Selberg and Oren Etzioni describe a metasearch engine as "the next level up in the information 'food chain'." The metasearch engine does not maintain a database, but submits a request simultaneously to several other web engines and/ or directories. The results may come back integrated into a single list, or sorted according to the engine that located them. Each item mayu or may not be given a relevance score, based primarily on the frequescy of your search terms within the document. You may select any item and immediatedly link with it.

MetaCrawler
This metasearch tool searches eight other search tools: Excite, Galaxy, InfoSeek, Inktomi, Lycos, OpenText, WebCrawler, and Yahoo. You may search for any word (like an OR command), all words (AND), or all words as a phrase. A minus sign preceeding a search term equates to NOT in Boolean searching, thus eliminating items containing that word. MetaCrawler allows each engine a limited response time, selected by the user. You may limit the search geographically or by organization type (e.g.,educational, commercial). Results are double-checked for availability, ranked for relevancy, and displayed beginning with the highest scoring item.
SavvySearch
This search tool displays its search form in any of several languages. It selects two to five databases from its list of (currently) 23, choosing on the basis of relevance to the topic and speed in responsse to recent inquiries. All eight of MetaCrawler's search engines are included. SavvySearch does not have the NOT equivalent. Up to 50, but no more, items may be accepted from any one of the underlying engines. Results are not checked or sorted. SavvySearch presents its search plan, showing the engines already selected as the best for your inquiry, followed by others you may try.
A Unified Search Engine for InTernet (USE IT!)
This is a recent addition to the metasearch group. Based in Italy, it features a number of international search tools, including business, news, and computer databases. You may choose either less than 20 or more than 20 items from the databases you select. Return time is set from one to five minutes. At present the best way to search is by entering a single word, usually in the language of the search tool. More complex search procedures are being developed.
The WWW Virtual Library
This is a distributed subject catalogue. See its Category Subtree, Library of Congress Classification (Experimental), Top Ten most popular Fields (Experimental), Statistics (Experimental), and Index. See also arrangement by service type, and other subject catalogues of network information .

The Comprehensive Search Tool

This tool reaches beyond the Internet to include familiar, commercially published indexes.

NLightn
This tool has the broadest scope. Using its Universal Index, NLightN not only sifts through the WWW (using Lycos, but simultaneously searchs 500 proprietary databases. These include
  • Business Periodicals Index
  • Education Index,
  • ERIC
  • Biography Master Index
  • Conference Papers Index
  • Encyclopedia of Associations
  • Newspaper and Periodical Absctracts

Many of these databases are available elsewhere, but for a fee. With NLightN, one may register and conduct a search without charge. Internet hits are then reachable free. For resrouces from the databases, NLightN often provides enough information so that one can locate the item. If not, or if one wants an abstract or full text, charges range upward from ten cents per citation.

Infobots and Personal Searchers

Information in this section was developed by Claire Carpenter, Technical Editor and Internet Trainer, Faculty Academic Computing and Technology Support (FACTS) Center at the University of KY.

Several personal robots and agents to bring you your own personal electronic newspaper have appeared in recent weeks (e.g., from InfoSeek, Excite, Yahoo, and Apple). New robots and personal agents to help you with Web searches you make regularly are no doubt in the offing.

The URL-Minder would like to be "your own personal Web robot." It can automate some aspects of keeping up with Web developments. When any Web resource you have registered changes, the URL-Minder will notify you by e-mail. If you register a search URL, it will let you know when something new has been found by your "standing" search request. See both the URL-Minder main site:

and "Clever Uses of the URL-Minder".

Poke! is a new "search assistant" or utility that works with Netscape to keep a record of places on the Web you have searched and explored. In essence, it seems to store your GO list from session to session instead of deleting it. Poke! stays right at hand and offers you ready access to several search engines. It's a promising tool, but still a bit buggy (save all open files before exploring Poke, since it may crash your computer).

"Agents of Change" What are these new "intelligent agents" anyway? For general information on what is on the horizon with artificial intelligence and software agents, see a recent article from Internet World. Don Norman, who is well known for his work in interface design (and his critiques of the poor design of VCRs!) is interviewed on the subject of artificial intelligence and "software agents" in a fascinating article, "Agents of Change," in the May 1996 issue of Internet World.

Popular Search Engines

Linking directly to Search Tools from Netscape using the Net Search button. This will take you to five popular search tools

The Search Engine Cheatsheet

If you are having problems understanding the Boolean operators in the various search engines, check out the cheat sheet.

Search Engine Comparison Information

You also can compare search engines at More Search Engines. You can access

Search Engines Plus

Making the Connections

This set of resources developed by Lillian Biermann Wehmeyer provides information on how to find teaching resrouces and discussion groups on the Net.

Virtual Libraries and Subject Indexes
of Particular Interest to Higher Ed

INFOMINE, a new and growing virtual library being constructed through the Library at the University of California, Riverside, includes over 5000 hand-picked links of potential interest to people in academe.

According to its opening blurb, "INFOMINE is intended for the introduction and use of Internet/Web resources of relevance to faculty, students, and research staff at the university level. It is being offered as a comprehensive showcase, virtual library and reference tool containing highly useful Internet/Web resources including databases, electronic journals, electronic books, bulletin boards, listservs, online library card catalogs, articles and directories of researchers, among many other types of information." There are currently eight different INFOMINES in various broad subject areas, each with special, very flexible and powerful search and browsing options. For more information, see the Introduction.

webCATS, Library Catalogs on the World-Wide Web

Increasingly, libraries whose catalogs had been available via telnet are now becoming reachable through the Web. According to the webCATS homepage, "This resource will link you to all library on-line public-access catalogues which have World Wide Web interfaces." Three different indexes offer you convenient options for moving to the libraries that interest you. The geographic index, for example, allows you to move from a continent to a county, while the type index allows you to start by type of library (public, law, college and university, etc.). A new search option should be added to this site soon. A link is also provided to the webCATS predecessor, HYTELNET, which continues to provide telnet access to hundreds of library catalogs worldwide, including over 1000 in the US.

Limited Area Search Engine

ARGOS: Limited Area Search of the Ancient World

A new venture in "peer-gathered" discipline-based database, ARGOS is a project coordinated by Prof. Tony Beavers, who teaches philosophy at the University of Evansville. He is working with colleagues in classical studies to prepare and maintain a searchable index of a few quality sites for students and researchers in classical (and medieval) materials. One of his associates is Prof. Ross Scaife, in Classics at UK. This pioneering effort, which claims to be the first LASE (Limited Area Search Engine), aims to help classicists find highly relevant Web materials without having to wade through dozens or hundreds of irrelevant links. Academics in other disciplines may want to undertake similar efforts to make searching on the Web more rewarding and fruitful.

Other Interesting Resources

Information in this section was developed by Claire Carpenter, Technical Editor and Internet Trainer, Faculty Academic Computing and Technology Support (FACTS) Center at the University of KY.

Several articles have appeared recently which provide pointers and ideas on guiding students (local or distant) as they learn searching skills and techniques.

To see how a research mentor can make the search process real for students, read "A Mentor/Research Model to Teach Library Skills: An Introduction to Database Searching" in the September 1996 issue of T.H.E. Journal, pp. 96-98. See the online copy, which should appear soon on the THE Journal website.

For guidelines on imparting to students a sequenced set of search strategies, read "Teaching Online Search Techniques Your Students Can Use," in the September 1996 issue of Syllabus. pp. 52-56. This article is unfortunately not available online, though you can read a short teaser.

The home page provides information on how to get a free subscription to the magazine, which provides excellent coverage of innovative computing in higher education.

UMDL The University of Michigan Digital Library (UMDL) is one of six libraries with special funding from the National Science Foundation as well as ARPA and NASA to explore and prototype the idea of a digital library, especially as a resource and a stimulus for inquiry-based learning. "UMDL is an experimental system designed to provide access to information related to Earth and Space Sciences. A wide variety of resources are available to patrons of the UMDL, including magazine and journal articles, encyclopedia entries, sounds, and data." The UMDL efforts are geared particularly toward providing library services over a digital network. To explore this resource, try out this link.

To peek behind the scenes, read the related article "Toward Inquiry-Based Education Through Interacting Software Agents". This somewhat technical article explains how UMDL is defining modular intelligent agents to work together in helping library users find what they want.

Glossary (Tennant, 1996, p38)

Adjacency
How close one search word should occur in relationto another search word. For example, one of the Open Text Web Index "Power Searching" options allows you to specify that one word should occur "near" to another or "followed by" another.
Boolean operators
The use of joining words includeing "and," "or," or "not" to describe the logical relationship of one or more search terms with one or more other search terms.
False drops
Items that are returned but are not what you were wanting, usually because the search terms used were present in the results but not in the context of your search. For example, terms with the same spelling but more than one meaning often cause false drops, as do terms that are roots of or parts of other terms.
Hits
Items returned from a search that match your search terms.
Results ranking
The process whereby search hits are ranked in order of computed relevance to your query. Results ranking attempts to put the resources most likely to match your information need at the top of the list of results.
Search terms
Words that describe the topic of your search.
Stemming
The process of chopping off, or truncating, search terms to improve matching with variations of that word; for example, removing "ing" and "ed" endings.

References

Wehmeyer, Lillian Biermann, (1996)
Internet Resources for course Development, Syllabus August 1996 pages 26-29.
Wehmeyer, Lillian Biermann, (1996)
Wide-Angle Searching on the World Wide Web, Syllabus June 1996 pages 34-37.
Tennant, Roy (1996)
The Best Tools for Searching on the Internet, Syllabus February 1996 pages 36-38.

This page was developed and is maintained by Rene M. Hales. Please send comments or suggestions to halesr@pop.uky.edu.

Last updated on October 2, 1996.