Just What Does it Mean to Search the Web?
"Isn't there a place I can go to search everything on the World Wide Web?"
No. Even the most comprehensive search engines at best might be able to cover 60% of what is out there. It changes and evolves too quickly to carefully documents its contents.
The Internet is huge, vast, and some might say, in horrific disarray. How do you find things in all of this? There is no one single index or tool to find everything and anything. However, there are some resources out there to make it a little easier. We will review some of the different types of search tools, show some examples, and then send you on your way to conduct your own web search.
If you do not find a good search type here, try the ever-growing list maintained at (where else?) Yahoo
or the extensive collection from CNET. If you think about the previous lesson, these are good starting points to finding other search sites!
Robots, Spiders, Worms, Wanderers, and Other Critters
The basis of a web search is a special computer program that someone writes that more or less tries to follow every hypertext link it can find, from the biggest sites to the most obscure. Such a program "critter" visits each web page, reports back to its home some information about what it found and keeps on going in robotic fashion. For more details, see WebCrawler's Robots, Wanderers, and Spiders..
The "home" is ideally a very large powerful (or series of powerful) computers with massive storage. What you are really searching when you use a web search site is the information that the wandering critter has sent home.
The field is pretty competitive with many sites boasting that they are the "biggest." What makes one better than another?
- Amount of information returned. Some searches will return only a list of links while others include a description or summary of what information is on the found page.
- Responsiveness of Search. Faster, of course, is preferred. Some sites will just return all of the "hits" it found, while others will return them in sets of 10 or 20.
- Relevance Rankings. The better sites will sort the returns in order of which is most relevant, usually based upon how many times it found your keywords in the entry for a particular site.
- Power Search. Many sites now offer an ability to construct complex search expresses such as:
Find all the sites where "Maricopa" occurs near "Community" but does NOT contain the word "Alan"
- Related Searches. Some sites offer an ability to construct a new search that matches the characteristics of one of the returned sites.
What Site to Use
After some experience, you may find that you prefer on search tool over the other. In many cases, you are looking web-wide for information related to a topic. In this case, you might want to use one of the Multi-Site Searches, Extensive Search, or Meta-Search Sites listed below.
Specific Web Searches
Sometimes you may be looking for something more specific, such as a fact. This is why the webhound asked you in the section on Scoping Your Search to try to identify the nature of your search.
These examples are meant to show you how a specific type of search engine might be more useful than one that tries to search everything and anything.
Energysearch provides Internet users with fast, accurate search results on global energy topics.
History On-Line allows you to search a database of British history, featuring Books in print published by academic history publishers in the UK; Journals and articles, many
with abstracts; and more.
Ditto.com provides a search engine to find pictures on web pages.
Geographic Searches; Information about Places
One World - Nations Online is organized by geographic region and is great for zeroing in on a country, information about that country, and links to more detailed information.
WhoWhere? offers a service to look up people on the Internet by name or organization
Download.com features an extensive search of major Internet software archives
1-800 Phone Numbers
AT&T offers a search of its Toll Free 800 directory
MedScape offers a access to information for medical consumers and practitioners, including a searchable database of medical articles
The Meta-Index for U.S. Legal Research, created by the Georgia State University College of Law, presents search forms that point to many US Federal Government law sources.
This is just a sampling of the vast types of Specific Search Tools on the Web. In this webhound's pocket is also a handy Swiss Web Knife.
These are places that have organized lists of forms and pointers to search engines at other sites. You might call this "Search the Searchers".
The BigHub lets users search multiple engines, Web directories and news databases simultaneously and receive one simple summary of results.
You can think of Ixquick as the metasearch equivalent of Google. It has the same kind of simple, clean interface, and a similar knack for turning up relevant listings in the first page or two of results. Ixquick ranks results based on the number of "top 10" rankings a site receives from the various search engines.
c|net's has one of the best organized set of links to searchable web sites.
ProFusion, powered by Intelliseek offers interesting ways of choosing which engines and directories to search ("Fastest 3," "Best 3" as well as a pick-and-choose approach).
Vivísimo has won lots of fans since it appeared in late 2000. The main attraction is an excellent interface that provides grouping and filtering of both results and the selection of engines to be queried. Vivisimo is unique for organizing search results with document clustering instead of long, tedious lists.
Multi-Site Web Searches
Of course, a preferred web search tool would be one site where you can enter some keywords, and it returns everything on the web that matches your query. The closest you will find are sites that will take your request and send it to the handful of top web site search engines available, Here are a few that work fairly well:
is "a no frills, 'when you don't have time to monkey around' metasearch engine which searches Google, AltaVista, MSN, Lycos, Yahoo!, Overture and others to search sites, newsgroups and image collections."
is a new multi-site search engine
MetaCrawler was one of the first multi-site search engines
the Mother of all Search engines
A webhound has to appreciate a dogpile! Certainly one of the more uniquely named, sites, DogPile very efficiently ranks and returns search results from multiple search engine sites.
Jeeves is one the most polite and amicable search sites, because you ask your question in the most simkle format, by entering your search as a question! Jeeves takes some logical guesses at where the answers might lie, and should he fail, he provides linked results from other web search engines.
Since we are the webhounds, we will give dogpile a try! For our search on the history of numbering systems in ancient Mayan and Egypt civilizations, we enter the keywords:
Egypt Mayan Civilization Number System
In a manner of minutes, we start to recieve results from more than 10 different search engines, a few that we had never heard of before!. Note that some returned 0 finds while others report 10,000. After some scrounging around we find a link under the results for the search at goto.com that lead us to the Ancient History links from the Mining Company Web Site finaly leading us to sites about the Mayan Calendar and Mayan Mathematics.
Now, if we try Ask Jeeves, we can enter a question like:
What are similarities about Egyptian and Mayan Number systems?
Many of the interpreted results are way off base, but under the rephrased question "Where can I find information on the Ancient Culture" Jeeves will return links about "ancient Egypt and Nubian Culture" as well as "Mayan Culture". Jeeves' requests to other search sites produce a few mnore that look promising.
Extensive Search Sites
Several new sites have arrived on the web scene in the last few "dog months" that are by themselves very useful for web searches because of the immense size of the database that is searched. Three of note are:
This is currently the one to beat: bigger, deeper, faster, more relevant and richer in features than all the other search engines. Google has a specialty in the way it returns your results, ranked by relevancy.
Although HotBot is now owned by Lycos (it began in 1996 as part of Wired magazine's online services), it remains an independent search engine with unique power search capabilities. Their interface is pretty easy to use, if you do not mind their loud colors.
They claim to be the "biggest" at Digital's Alta Vista search. And they might be correct! The response is very quick and extensive.
Excite offers a normal keyword web search as well as a "concept-based" search that not only finds the words that you enter but others that are closely related. The also offer a searchable collection of reviewed sites and the ability to search the archives of UseNet Newsgroups
It is nice to get extensive results, but sometimes they are too extensive. That's where it pays to investigate if Power Search options are available. Most of these sites will have explanations listed under Help on their page.
Power searches allow you to:
- Find exact phrase matches. If I am looking for information about mountains in New Mexico, I don't want every document that contains the keywords "New" and "Mexico"; just the ones that contain them adjacent to each other. The most typical way to do this is to enter the phrase in quotes in the keyword field:
"New Mexico" Mountains
- Logical Expressions. Such "boolean" searches might allow me to search for all documents that contain Egypt and Numbers but not Population:
Egypt AND Numbers NOT Population
Or if we were searching wide, we might put:
(Egypt OR Mayan) AND Numbers
which would hopefully return all documents that contain "Numbers" and either "Egypt" or "Mayan"
(You should check each site for their rules on these types of searches. It does vary from site to site)
- Proximity. Sometimes you can specify to find only documents that contain words that are "close" in spacing to each other. This might work to return only web pages that contain the word "Egypt" that is within, say, 5 words of the word "Numbers"
Here's what to do...
Now it is time to conduct your own webhounding with search engines. We again have set up a worksheet for your convenience.
Look at where you are...
- Can you explain to a friend or colleague how a web "robot" works?
- As you explored different search engines, did you find the same results for the same search terms?
- Which engine seemed to produce the information in the useful format? Which one(s) were the quickest responding?
- How do you do a "power" search?
The next step...
You have reached the end of the webhound's tour. If you feel like a true webhound, sit back now and howl at the moon! Maybe you are ready to be a webhound leader and teach others what you have learned. Or, you can scoot back to the webhound's home to review another section.