InfraSearch is a simple and free search engine that distributes queries to remote providers. It allows providers to answer queries the way they want to and to return responses in a way that makes sense to them.
The providers may choose to respond to queries with real-time or dynamic data.
Becoming a provider on the InfraSearch network is as easy as:
1. InfraSearch queries are matched with provider registrations and are sent only to matching providers. An InfraSearch registration for a site that contains all kinds of rhubarb pie recipes looks like:
<?xml version='1.0'?> <register> <schema>http://www.infrasearch.com/food/recipes/</schema> <predicate> <and> <or> <type>dessert</type> <type>continental</type> </or> <ingredients>rhubarb strawberries bananas cherries</ingredients> </and> </predicate> </register>
The schema identifies the type of content that is shared by the node, while the predicate summarizes the content. The predicate is required to be in conjunctive-normal form, hopefully with not too many clauses. Each or-clause, however, can be rather large, so as to accomodate larger dictionaries.
2. To find out what can be made from strawberries and rhubarb, one should send a query like:
<?xml version='1.0'?> <query> <schema>http://www.infrasearch.com/food/recipes/</schema> <type>dessert</type> <ingredients>strawberries rhubarb</ingredients> </query>
Here are some examples:
Remember that the position of your response in the search result page is determined by various ranking heuristics. They take into account the time your node took to respond to the query, the size of the response, and how the person searching for the information reacted to the information in the response. Therefore, it is always a good idea to make sure you node answers queries quickly with small responses and that the information contained in the responses is relevant to the original query. Best of luck!