A search engine is a set of programs used to search for information within a specific realm and collate that information in a database. People often use this term in reference to Internet search engines, which are specifically designed to search the Internet, but they can also be devised for offline content, such as a library catalog, the contents of a personal hard drive, or a catalog of museum collections. These programs help people to organize and display information in a way that makes it readily accessible.
There are three aspects to a search engine: crawling, indexing, and searching. When one crawls, it looks for new content that was not present during the last crawl, including updates to files and web pages. Then, the programs index the information, pulling out specific keywords to categorize it. On the Internet, for example, indexing relies heavily on keywords in web pages and meta tags that provide information about the page.
Once the information has been indexed, information about how to access it is stored in a database. Some programs also store or “cache” information to make it easier to retrieve. When someone searches the database, it spits out results ranked by relevance. On wiseGEEK, for example, a search for “giraffe” will turn up articles related to giraffes, including, of course, "What is a Giraffe?"
While all search engines work in similar ways, their usefulness can be quite varied. They rely heavily on complex algorithms to rank the relevance of their search results, especially those for common keywords. Users tend to gravitate to ones that return results they like, with Internet sites like Google®, Yahoo!®, and Bing® warring for users with various features that are designed to make their searches more appealing and more relevant.
Google® has managed to become so associated with the process of conducting a web search that this process is often colloquially known as “googling.” The company is actually not very pleased with this, as it fears that lower case use of its name, along with generic use, could contribute to trademark dilution.
Many Internet search engines are smart enough to learn from their users, incorporating user activities into their relevancy rankings. They also rely on information like links from other pages and a site's reputation to rank search results, all within the fraction of a section. Adept users can sometimes manipulate search results, but many programs change and evolve to help combat this practice.