Comparitive Searching

I’m not a big fan of meta search engines, but one does standout and find its way into my regular tool-kit. One of the problems with Internet searching is building the correct string of search terms for a particular search engine.

Search Engines

I use clustering search engines and directories to help build the most specific search statement, but I’m still left with the problem of knowing if I have found the best search engine to use for that statement. The large search engines, Google, MSN (Live Search), and Yahoo offer the largest databases, but it takes more than typing in a person’s name to get relevant results. Long search statements must be structured properly for each search engine. Trying to create the best search statement for Yahoo or Live Search takes some effort starting with a hunt for the list of operators and the syntax documentation. You will notice that a comparison with Google that the search operators and syntax are somewhat different.

Operators

A meta-search that searches these three search engines gives-up some operators. Such a rudimentary meta-search site would not be of much use; unless it allowed you to compare the results from each search engine. It would also be useless if it didn’t allow the NOT, PHRASE, and MUST CONTAIN operators in a form that is recognised by all the search engines. Jux2 provides this functionality.

Jux2.com

Jux2.com allows you to enter a search statement and compare the results from the three databases. The plus sign is the MUST CONTAIN operator for both single terms and phrases. The PHRASE operator uses quotation marks. The NOT operator uses the minus sign.

I normally build the search statement using a person’s name or a company’s name as a phrase, then I add the NOT operator to eliminate irrelevant material. In this type of searching the NOT operator is more valuable than the MUST CONTAIN operator, which will limit your results to terms you know are related to the subject. The unknown, but relevant, information may be eliminated from the results if the MUST CONTAIN terms are not present. In essence, you will find only what you already know about. Use the NOT operator to eliminate material that you know to be irrelevant. This allows information you don’t know about, that may be relevant, to rise to the top.

For example, the following search statement seems to work in all three search engines and Jux2:

+”richard mceachin” -stockton -bioinfo-echostar -Coffeyville -Catcher -”David J. States” -brentwood -bioinformatics -TYBEE -Genetics -genetic -oakland -clayton -ocean -shores -Deckerville -Fairfax -rootsweb -oakley

Be careful with the minus sign — it must not have a space after it.

In Jux2 we can see that this is too specific for MSN Live; Google returns about 770 results; and about 470 hits appear in Yahoo. Jux2 will show you what results are exclusive to each search engine.

My experience indicates that the major search engines will duplicate 5% or less of the results on the first page with any give search string. Using Jux2 demonstrates the best search statement for each search engine. To do this, you adjust the search string when it provides poor results for one engine. As you do comparisons, the best search strings are copied into Ditto or your text editor. Go to one of the search engines, use your “best” search statement for that search engine, then refine its results using the operators and arguments not used by Jux2 and by adding search terms.

This may seem like an involved process, but doing this will help you learn the operators and syntax of these massive search engines, and teach you to build the search statement for each one in the most efficient manner I am aware of, by hands-on experience. Your goal is to build search statements that deliver the most relevant results in the first few pages.

0 Responses to “Comparitive Searching”


  1. No Comments

Leave a Reply