The Aristotle Semantic Network Technology

A novel way to represent knowledge using extended semantic networks as underlying model is presented. This approach extends the semantic network idea from its frequent application in presenting human-readable data correlations into a basis for a general computerized data modelling technology. The method was successfully applied using biological data from the Swiss-Prot and TrEMBL sections of the UniProt knowledge base to facilitate high throughput data mining applications. It is also shown that the method can be effectively used to automatically generate the relational implementation of a persistent warehousing system containing data from heterogeneous resources. The approach is extended to an object-oriented level, including the automated creation of object-relational mappings between objects held as data structures in an application framework and the underlying relational data warehouse component. The technology is used as persistency level of parts of the UniProt web site (see http://www.ebi.uniprot.org).