All About Search Engines

Types | Selecting the right one | How they work

Search Engines

are programs that let you do keyword searches for information on the Internet. They create indexes of databases or Internet sites based on the titles of files, key words, or full text. They provide an interface allowing you to type in your search, then return with a list of search results in hypertext, allowing you to click on any item in the list to get the actual file. If the file you select is not what you want, use the Back button on your browser to return to the list of search results and make another selection.

Types of Search Engines

Subject Directories

are usually human-compiled web guides that group web sites together under similar categories and are useful for identifying general information.

Yahoo!'s Directory is a popular hierarchical subject directory.

Web Indexes (General Search Engines)

search all the contents of a web site using software programs called spiders or robots. These robots search the Internet, analyzing millions of web pages, news group postings, and new web, gopher and FTP sites, indexing all words. Robots search for documents within web sites that match your search terms. After entering a keyword search, the Indexing Robot assigns a percentage rating for relevancy of each search return.

WebCrawler is a popular web indexing search engine.

Specialty Search Engines

index or categorize certain types of items like news articles, discussion groups, FAQs (frequently asked questions), or Web Forums. Popular ones include:

Google Groups Post and read comments in Usenet discussion forums.

Google News

Usenet FAQ Archive Search

Metasearch Engines

run your search on multiple search engines and display results lists together.

Dogpile

MetaCrawler

Natural Language Searching

engines use Natural Language Processing (NLP) to match query concepts phrased as questions. They statistically and linguistically analyze your question and act as question answering system. Searches are phrased as questions such as "What is the population of China?"

AltaVista

Ask

Select the Right Search Engine

Each search engine has slightly different rules for searching.

Get to know several search engines well by experimenting.

Visit the Help or Tips area and read the screens.

The best matches are usually displayed at the top of the results list.

Search engine features to consider and compare:

Is the response time quick or slow?

How easy or difficult was it to figure out how to search?

Were there useful help screens to guide you in formulating your search?

Did you get relevant, current results?

Was the amount of detail displayed with the results adequate?

Was the order in which the results were displayed evident and/or logical?

Were the results list display features helpful?

How Search Software Works

Understanding how search engines work will increase your effectiveness when searching.

Different search engines use different types of algorithms to search, so results differ widely.

Various search engines cover different areas of the Internet. There is no comprehensive way to search the Internet in one search.

Search engines use statistics and probability to match documents to your query and rank their degree of relevance. The more often your terms appear in the document, the more relevant that document is considered.

The length of the document can affect its relevancy ranking. When terms appear at the beginning of a document, many search engines assign that document more relevancy.

Relevancy is also assigned by how close multiple terms appear together.

Most search engines provide automatic stemming, truncation, and synonyms. They also have stop words, which means they eliminate very common terms from your search.

Results frequently include exact and fuzzy matches. Fuzzy matches may have two of your terms, but not the third. These partial matches broaden your search.