Our work on multidatabases was the jumping off point for most of the early research undertaken at the ADSL at the University of Manitoba. To date, we have developed a tool capable of extracting semantic information from various well structured data sources (they must be ODBC compliant). The tool does much of its work automatically but it requires user interaction to ensure the proper semantics are attached to new terms that appear during the semantic extraction process. This will be required for any practical tool and we have shown that for realistic real-world examples constraining the ontology is both good and necessary. This work is ongoing but it is now found under other headings including Bioinformatics, Datawarehouses, and Mobile Computing.