Zen & The Art of Internet Searching
By Margaret Carol Fine
"You see things vacationing on a motorcycle in a way that is completely different from any other. . . . In a car you're always in a compartment . . . (with) a cycle the frame is gone. You're completely in contact with it all. You're in the scene."
Zen & The Art of Motorcycle Maintenance
ROBERT PIRSIG wrote about his motorcycle travels, and the magic of seeing America from the open road, when he published his literary sensation in 1974. Today when searching for information, we often find the Internet has a boundless quality that allows us to travel on an open road through gigabytes of electronic data.
In fact, Internet searching has become an art. While search engine algorithms are designed to retrieve high volumes of data in a speedy manner, users must create intelligent queries to get useful data results. Available search languages give users powerful tools to extract data, including key word, wildcard, numeric, Boolean, concept, fuzzy logic, and other types of searches.
Depending on the search engine, users can create a single search request, or multiple levels of searches within a single request or a "hit" list. Meta search engines like Dogpile and MetaCrawler also query several Web search engines at once.
Here are some tips to help improve your searching results. Examples of major search engines are given to illustrate searching features and capabilities, but this article is not exhaustive.
Start with basic strategies and go straight to the source. If that fails, use more advanced search options to narrow searches to retrieve data.
Because "key word" searches use one or more words to help a user find a topic, they are often a good starting place to locate relevant data in a Web-based repository. Most search engines offer key word searches; for example, Google is known for its high relevancy rating in returning important, (i.e. relevant) hits.
Refining Your Search
If a simple key word search doesn't get leads or blaze a trail, the challenge becomes how to start refining a search to get results.
Try using convenient search engine features to narrow a search request such as related searches, clustering, word stemming, data ranges, case sensitivity, among others. For instance, clustering allows only one Web page per site to show in the search results such as on Google and HotBot.
Besides search results, Alta Vista, HotBot, Go (Infoseek) and other search engines list related search topics. Using an exact phrase statement, usually placed in quotes like "motorcycle maintenance," tells a search engine to find that precise statement.
Field searching is also available to find a subject. For instance, a user can field search on Northern Light by title, URL, company, pub, text, or ticker.
Truncation or word stemming can take a word like "travel" and give you derivations of that word like travels, traveler, traveled, and traveling. Sometimes word stemming also refers to wildcard searching, or rather looking for wildcard characters related to a word stem. Users can enable word stemming to get derivations of a word on HotBot Advanced, MSN Advanced, and other search engines.
Further, AltaVista Advanced, HotBot Advanced, MSN Advanced, and Northern Light accept data ranges to narrow a search.
Also, check to see if capitalization matters in creating a search request; search engines can be case sensitive. Some search engines like Go (Infoseek) allow a user to search within the results of a single search request.
Boolean searches (using the Boolean operators "AND", "OR", "NOT", "NEAR") use specific logical sequences of key words and operators to restrict a search to retrieve a certain set of data. Boolean operators help narrow and refine a search request, but not every search engine provides for Boolean searching.
You may have to poke around on a search page to find where to create a Boolean search request: placement can differ among search engines. For instance, the box labeled "Boolean query" is the one to use on the advanced search option for AltaVista. In addition, search engines can be case sensitive for Boolean operators and the use of these operators can vary.
Excite requires Boolean operators in upper case letters. Generally, most search engines use the Boolean "AND" or "OR" operators, but the proximity operator "NEAR" can vary from one search engine to another. The "NEAR" operator finds terms in close proximity to other words. For instance, "NEAR" means within 10 words of another term on an AltaVista Advanced.
Boolean searches combined with other search features and capabilities can further refine a search for more precise results. For instance, a Boolean search and an exact phrase search such as Zen OR "motorcycle maintenance" will retrieve only data containing either the term "Zen" or the phrase "motorcycle maintenance," or data containing both the term and phrase together.
Likewise, wildcards can expand the scope of a search request by allowing for derivations of a word, instead of just the word itself. The use of a proximity operator such as "NEAR" may find words within 10 words in proximity to a word. A "NOT" operator can exclude all references to a word and its derivation.
The following example shows a narrow search using multiple levels of search language criteria as described above: ((motorcycle OR maintenance) NEAR Zen) AND NOT auto*
The use of parentheses groups search queries into more complex queries, and are called nesting. Nested queries are evaluated by working outward from the innermost group of parenthesis, and applying the search operators and criteria to determine the final result set. The search results for this example will return a subset of data returned by the first search example, Zen OR "motorcycle maintenance."
Concept searching involves using a massive dictionary of the English language to find synonyms and related words. Search engines like Excite are designed to identify several words and phrases with a particular concept using statistical analysis.
A concept search for the word "motorcycle" might pull up the words, "moped" or "bike," or even broader related words like "automobile," "car," or "vehicle." Generally, concept searching is useful to focus topics when you're lost and can't find your exact destination until the general area is located.
Fuzzy logic searching finds matches based on searching for degrees of certainty as opposed to restrictive Boolean logic. Fuzzy logic finds sound alike versions of words such as incomplete or misspelled words like "motercyler."
In addition to enhancing the accuracy of major Internet search engines, fuzzy logic is particularly important for finding data in Web-based repositories that contain scanned and optical character recognition (OCR) data.
Like riding on the open road, the Internet gives users a sense that they can travel through electronic data and find whatever they need. Today there are many search engine features and capabilities designed to get precise search results. Depending on how well a user can navigate search engines, he or she may have a good trip while seeking that information or knowledge. Like Pirsig once said, "A motorcycle functions entirely in accordance with the laws of reason, and a study of motorcycle maintenance is really a miniature study of the act of rationality itself."
Attorney Margaret Fine resides in Portland, Ore.