User Tools

Site Tools


podoc:search

This is an old revision of the document!


SEARCH ENGINE BASICS

There are three categories on which Photo Organizer's search engine can perform queries. These categories were introduced to improve the search performance and to allow customized display of the results.

The search engine uses stemming to automatically include related variations of your keywords. If you search for update, you will also find updated, updates, updating and so on. Every word in a search query is stemmed unless quotation marks are used.

The search engine will ignore certain, administrator customizable keywords. By default the system will ignore all single letter words as well as the most common two letter words. The engine will warn you if some of your keywords were ignored while providing a possibility to force the engine to search these “forbidden” keywords by using quotation marks.

The general rule is that quotation marks around your search terms will tell the search engine to skip all preprocessing and feed them directly to the query.

The syntax of a search query is a list of keywords and operators separated by spaces or tabs where operators must be in upper case (otherwise they are considered keywords):

“query :- OPERATOR keyword OPERATOR query”
“query :- keyword”

The three basic search operators are:

  • AND - simple logic AND - this is the default operator, which is used in case no operator is specified between two keywords.
  • NOT - simple logic NOT
  • OR - simple logic OR

During a general photo search, the search engine will search on the photo's title, author, caption, keywords, shooting location and unique photo identifier. A simple folder & album search will search only on folder+album names, while a simple user search will execute the query on the users' first, last and username.

Some examples for the photo category:

While the general search engine will search for all search keywords on all fields, the advanced search engine will match search keywords with dedicated fields only, increasing the engine's flexibility, response time and accuracy.

Currently there are 6 separate fields that basically perform independent search queries and each of them return an independent set of pictures. The advanced search engine introduces two additional operators to manipulate these sets.

  • UNION - will create a union between the picture sets - will add picture sets together.
  • INTERSECT - will create an intersection of the picture sets - will return the comon pictures only.

We will illustrate how these opeators work through an example. Say, we search on Title, Author and Caption simultanously. The operator for tht Title will be set to UNION, while the operator for the Author will be INTERSECT and the operator for the Caption will be UNION. First the search engine will perform a search on the Title and the resulted set will be united with the NULL set therefore all photos returned by the first query will enter into the temporary set of photos. Secondly the search engine will perform a search on the Author fields, and the resulting set of photos will be intersected with the existing temporary set of photos, therefore a new temporary set is created that contain common photos of the two sets. Finally the search engine will search on the Caption field, and the resulting set will be merged with the current temporary set. After all queries are executed the temporary set will be returned as the result of the Advanced Search query.

SEARCH RESULTS

Depending on the category, search result will be displayed in different ways. The folders+albums and users will be listed in alphabetical order while the photos will be presented in a similar way as for folders+albums. Photos can be selected the same way as for folders and they can be linked to albums. This allows you to store your search results (or a subset of your search results) into albums, for further reference.

podoc/search.1164218092.txt.gz · Last modified: 2006/11/22 17:54 by Luud