Put simply, the search scans documents in the site for the terms you input, returning documents containing all those terms. If you search for..
you will get back a list of all documents containing the word “cakes”. If you search for..
you will get back a list of documents containing the word “cakes” AND the word “standard”. In other words, the search operates in boolean AND mode. You can narrow down a search by simply entering more and more relevant query terms. “stop-words” (such as “of”, “the”, etc), are removed from the search query, and are not taken into consideration.
The search scores documents by their relevance to your input query. The scoring mechanism is weighted towards terms that appear earlier in your search string, for instance, a search for
rockets candles cakes
would return exactly the same documents as a search for..
cakes candles rockets
but the results would be returned in a different order, The first with documents more relevant to “rockets”, then “candles”, etc, and the second with documents that featured more occurrences of the word “cakes”. This usually gets the most relevant document to the top of the list, useful in a big search with multiple terms.
Titles and Filenames:
Other aspects of a document are taken into account too.
In your search for..
any document containing either “cat3” or “candles” in its title, or its filename will receive a bonus.If the entire phrase “cat3 candles” is there, a further bonus is awarded. Another way of getting the most relevant files to the top of your search results.
As well as the individual query terms, the search considers the query phrase as a whole, and scores documents containing “phrases” more highly. if you searched for..
all the documents containing both words would be returned, but any documents containing the phrase “cat3 fireworks “, would be ranked higher. Any “stop-words” (“and”, “of”, etc) in the original query are taken into consideration when looking for “phrases”.