LuMriX is a search engine that exploits XML and XML Topic Maps. In contrast to other retrieval methods, it does not relate single items to resources, but combines given items into meaningful associations (concepts), which are in turn linked to resources. XML Topic Maps allow an intelligent mapping of relations between terms and pages. The meaning of the query is captured by transverse joint relations between the search items. LuMriX is also able to auto-extend its thesaurus and create new relations between failed search items and information resources. It is completely implemented in Java. It can consist of many individual distributed LuMriX servers, which communicate with each other by distributed algorithms. Standardized interfaces such as TCP, SOAP, HTTP, XML, and XTM allow simple utilization and maintenance by other applications.