MasterKey is a metasearch system. It is a single point of access to a given number of individual databases.
MK2 (mk2.indexdata.com) is a public demo that exhibits various features of the platform.
You can specify simple phrases, or complex boolean queries, and you can limit searches to certain fields if desired. The following search fields are available:
Examples:
hello kitty
ti="hello kitty" and date=1999
ti=hello kitty and (au=smith or au=jones)
Quotes and parenthesis are optional, but can be used to disambiguate complex queries, or queries that contain words like 'and' which can be mistaken for operators.
If your result set is large, it will be truncated by the metasearch engine. Try to make queries as precise as possible. If your result set is large, you can use the facets on the left to narrow the search until you find what you are looking for (see later).
You can sort results by Relevance, Title, or Date (descending or ascending). Relevance ranking takes place on the client side according to the popular TF-IDF algorithm.
You can specify your search criteria further by using the facets on the left. For instance, to limit results to a certain media type or to a certain source database, select the items of interest from the Medium or Source facets.
The subject facet is represented as a 'tag cloud', meaning that a term in a larger font indicates a subject that appears more frequently in the current set of results.
Pazpar2 is a high-performance metasearch system, designed to make results available to the user from the target databases as quickly as possible. That means that as soon as results become available, they are displayed. However, the performance of individual back-end databases are outside of our control. What this means is that if one database is slow to respond, the system will keep waiting until all results have been returned. However, during this time, the user can browse the available results, change the sort order, start new searches, and generally use the system normally.
This metasearch interface is based on Pazpar2 from Index Data.