How Internet Search Engines Work

Once the spiders have completed the task of finding information on Web pages (and we should note that this is a task that is never actually completed -- the constantly changing nature of the Web means that the spiders are always crawling), the search engine must store the information in a way that makes it useful. There are two key components involved in making the gathered data accessible to users:

In the simplest case, a search engine could just store the word and the URL where it was found. In reality, this would make for an engine of limited use, since there would be no way of telling whether the word was used in an important or a trivial way on the page, whether the word was used once or many times or whether the page contained links to other pages containing the word. In other words, there would be no way of building the ranking list that tries to present the most useful pages at the top of the list of search results.

To make for more useful results, most search engines store more than just the word and URL. An engine might store the number of times that the word appears on a page. The engine might assign a weight to each entry, with increasing values assigned to words as they appear near the top of the document, in sub-headings, in links, in the meta tags or in the title of the page. Each commercial search engine has a different formula for assigning weight to the words in its index. This is one of the reasons that a search for the same word on different search engines will produce different lists, with the pages presented in different orders.

Regardless of the precise combination of additional pieces of information stored by a search engine, the data will be encoded to save storage space. For example, the original Google paper describes using 2 bytes, of 8 bits each, to store information on weighting -- whether the word was capitalized, its font size, position, and other information to help in ranking the hit. Each factor might take up 2 or 3 bits within the 2-byte grouping (8 bits = 1 byte). As a result, a great deal of information can be stored in a very compact form. After the information is compacted, it's ready for indexing.

An index has a single purpose: It allows information to be found as quickly as possible. There are quite a few ways for an index to be built, but one of the most effective ways is to build a hash table. In hashing, a formula is applied to attach a numerical value to each word. The formula is designed to evenly distribute the entries across a predetermined number of divisions. This numerical distribution is different from the distribution of words across the alphabet, and that is the key to a hash table's effectiveness.

In English, there are some letters that begin many words, while others begin fewer. You'll find, for example, that the "M" section of the dictionary is much thicker than the "X" section. This inequity means that finding a word beginning with a very "popular" letter could take much longer than finding a word that begins with a less popular one. Hashing evens out the difference, and reduces the average time it takes to find an entry. It also separates the index from the actual entry. The hash table contains the hashed number along with a pointer to the actual data, which can be sorted in whichever way allows it to be stored most efficiently. The combination of efficient indexing and effective storage makes it possible to get results quickly, even when the user creates a complicated search.