Effective Retrieval Performed by DIMES with the Application of Lucene

Earth science metadata search engines dealing with huge volumes of Earth science metadata need an effective method to respond to queries and retrieve metadata very quickly. Basically, Earth science metadata includes three types of information: descriptive text, temporal range, and spatial range. So, except for the general text retrieval, Earth science metadata retrieval requires a temporal range query and a spatial range query, which are not widely needed by other kinds of search engines.

DIMES, a distributed metadata server, is testing use of Lucene to tackle the metadata search problem of Earth science metadata. The vantage points of Lucene are that it is not only a high-performance text search index, but also has a special query type called range query which is especially fit for temporal range and spatial range queries of Earth science metadata once the temporal and spatial information are converted to strings with alphabetic orders. This paper introduces how DIMES applies the Lucene range query for the above temporal and spatial range search.