Criteria and Ratings of Search Engines

Overview

Search engines are databases that take Web documents, consisting of words and images, and create indexes for them. In order to make this possible, and to build databases, software robots survey the Web. The documents are then retrieved and indexed. The indexes can then be searched when the consumer makes a request for information.

When the consumer enters information into the search engine’s website, the words typed will be compared with the words in the search engine’s keyword indices.  The search engine will then return the best-matched websites to the consumer – in theory!

Making Search Engines Work for You

There are two ways that a consumer can use a search engine for an inquiry. He can search for text using either a keyword or context search. Keyword searching is used much more frequently because it is difficult to determine what the concept is for each individual Web page.

Keyword Searching

As said before, keyword searching is more common than context searching due to the fact that most search engine companies find it to be an easier means to retrieve information from websites. The search engine company pre-determines the keywords for individual websites so that when a query is entered at the search engine’s website, the significant word is flagged and displayed to the consumer. The words chosen by the companies are normally the words that are deemed to be important.  Other important words are those that are repeated periodically throughout the website.

Concept-Based Searching

Concept-based searching is much more difficult for the search engine companies to determine because instead of merely looking at the text’s words, the company must read the words and give them meaning. So, when doing a concept-based search, the consumer would be looking for what a certain document or website is about, instead of what it contains textually.  In other words, he will find the subject of a document even if the words typed in the search engine’s query box do not match those, which were retrieved. According to The Spider’s Apprentice, concept-based searching is often called “clustering”.  This means that the “words are examined in relation to other words found nearby.”

Criteria for Ranking a Search Engine

Linkage

Quantity

Most search engines count linkage as a ranking criterion. This means that when the search engine reaches a link on the Web, it checks to see how many times that link has been “hit” or to see if it is a member of its database of sites. If the link is already a member of the database, the engine then checks to see how many other links are pointing to it. The more links the site has, the higher it is ranked. Many large search engines such as, Google and Excite take the quantity of linkage into account.

Quality

Linkage is an important factor in weighing search engines. Not only does the amount of links pointing to a specific document in a database matter, but the quality of those websites is taken into account as well. So, if looking at a specific link, the consumer sees that another well-known, good site is pointing to that link, then it is easy to conclude that the database must also be good. This will in turn raise the engines ranking.

Out-Going Links

To determine the consistency of a site, the consumer can also look to see if the individual document not only has links pointing to it, but whether it points to other links as well.  If the links are going both in and out, it causes an intricate “Web” of information. The more detailed the engine, the more webbing can occur to and from popular, high quality links.

Word Indexing Capacity

Another criterion of good search engines is determined by the amount of word indexing done on each page of a document.  Most search engines index every word on every page.  These engines are more than likely going to be the larger companies. Other search engines index only part of a document.  When they do this, they are looking at the title, headings, subheadings, the first 20 lines of the text, and hyperlinks to other sites.

Although partial-text indexing was more common only a few years ago, full-text indexing is generally seen more often in today’s search engines. Full-text indexing looks at all words in the text and picks up on all of them except for the common fillers, such as “the”, “and”, “or”, and “www.” Some search engines discriminate upper case and lower case, while other search engines do not.  These facts go to show that some search engines are more detailed than others.  In fact, Alta Vista claims that it looks at all words to do indexing, even including the “ands”, “ors”, and “buts”.

Frequency

Another way that search engines can be judged is by their frequency.  The way an engine can be named frequent is if in their indexing, they are able to find repeat words to show whether a document is relevant to a query made. For example, if the consumer wanted to find out about ballet, then the engine would look for that word in a document.  The more times that word is found throughout a document, the more likely it will be useful to the consumer. However important frequency is, some more advanced search engines also add placement of the words into their searches.

Examples of Good Search Engines

Alta Vista

·        Is a keyword search engine
·        Searches the Web and Usenet
·        Can be refined by using Boolean “AND”, “OR” and “NOT”. 
·        Can also be refined by using the proximal locator “NEAR”
·        Ranks according to how many and how close together the search terms are
·        Results are presented as the first few lines of a document|
·        Possibly the headline

·        Possibly the first line of the paragraph

Fast

·        Is a keyword search engine
·        Searches the Web, mobile web, FTP files, multi-media, MP3s
·        Will search any foreign language
·        Can be refined by using Boolean phrases
·        Results are clean and fast
·        Rated to be one of the overall bests

Google

·        Is a keyword search engine
·        Searches the Web, news, stock quotes, special searches on Government documents, and PDF documents
·        Can be refined by using any Boolean phrases
·        Looks at both quality and quantity of links pointing to and from links
·        It is clean and simple with high relevancy results, which make it one of the best search engines today.

Related Links

Search Engines and Off-Page Ranking Criteria

A to Z Web Site Promotion

Search Engine Optimization Promotion Tools

The Spider’s Apprentice: How to Use Web Search Engines

A helpful guide to search engines - How search engines work

Locating and evaluating Web resources