|
ACADEMIC COMPUTING and COMMUNICATIONS CENTER | ||||||||
|
|
|
|
|
|
|
|
|
|
| ||||||||||||||||||||||||||||||||
Fields & Queries | ||||||||||||||||||||||||||||||||
| What Fields Can You Search On? | ||||||||||||||||||||||||||||||||
|
Netscape Catalog creates a summary of each document; it is this summary that is actually indexed and returned by the search engine. Each summary consists of several named fields and associated values. All of the values are combined and indexed, and in addition, some of the fields are indexed separately. In order to formulate intelligent queries, it's important to understand where the values come from, and which ones are indexed. For example, the title field is derived from the <title> HTML element, and it is indexed. Therefore one could execute the query: find all documents where the word finance occurs in the title field. (The exact query language is different, of course, and will be described in another section.) The author field is also indexed, but it is derived from the <meta> tag. That means that the HTML author must insert this into the <head> element of the HTML file:
<meta name=author content="Paul Dirac">
After that, it would be possible to execute the query: find all documents
where the word Paul occurs in the author field.
The partial-text field is not indexed; it is generated by the approximately first 1kB of text from the HTML document. If the word finance is contained in the partial-text extraction, it would be possible to execute the query: find all documents with the word finance in any field. But note, it would not be possible to restrict the search to the partial-text field, because this field is not indexed separately. Some of the fields you might find useful are listed in the following table. Some of these fields are generated automatically, but others must be added explicitly with the <meta> tag. Some are indexed and can therefore be searched separately, while others are not specially indexed and can only be searched as part of the overall document index.
|
||||||||||||||||||||||||||||||||
| Construction of Queries | ||||||||||||||||||||||||||||||||
|
The Verity search engine will accept a wide variety of queries. It can search for documents based on word stem, exact word, wildcards, relevance ranking, and so forth. For full details, please consult the Verity Site.
This is of main interest if you want to formulate your own advanced
queries. Otherwise, the cgi scripts at UIC will translate simple
words into simple queries. But if you understand a little of the
verity syntax, it will be clearer later how the web form input actually
gets translated to a real search.
(In most searches, stemmed variations of the search words count
as a match. So if you search for This table contains a small sampling of the most important search operators.
|
||||||||||||||||||||||||||||||||
| Web Search Forms | Previous: 3 What's Indexed | Next: 5 Forms |
| 2002-6-29 wwwtech@uic.edu |
|
