How to Set-up RDF Converter for you local machine The RDFConverter tool relies on the Jena API, at (http://sourceforge.net/projects/jena/files/Jena/Jena-2.6.4/), so all of the jars in the Jena distribution need to be on the build path for this project (as does some version of a mysql-connector jar). The packages of interest in the SDD project folder are conversion and dao. The conversion package holds RDFConverter.java and a sub-package test with a unit test demonstrating the use thereof. The dao package contains a couple of data-access java files for looking at two different databases: one database contains a single table that maps singular names to plural names. The other maps filenames (the output files of Hong's algorithms) to taxa. These are the only databases used for the whole "SDD" project (the project in which I'm doing all my work on SDD, RDF and matrices). To set-up these databases on your machine: First, there are sql scripts for loading these ("fnav19_filename2taxon_4boston20110412 0909.sql" and "singularpluralorgannames.sql"). Run these with mysql to get databases/tables installed. Next, make sure some mysql user account you have access to has the proper privaleges for accessing these tables (select privelages are enough). For instance, I created a user called "biosemantics" identified by password "stimpy" that has all privaleges for only these tables from any location. You could also just use your root mysql user. Once those tables are set up, find the properties file called "database.properties" in the dao package. Change the user and password properties to match your mysql user and password. Save the file. Input and Output: You need to tell the object that parses annotations the path to where those annotations are located. This can be done in "description.properties" in the conversion package, by changing "input.path" to point to the path on your file system where the annotation XML files reside. You don't need to worry about the other properties in description.properties. Very briefly, the way I begin generating either SDD or XML is by first creating a DescriptionParser object for a given taxon name and rank (corresponding to "cirsium" and TaxonRank.GENUS in TestRDFConverter.java). This is where the filename2taxon database is used - a filename is looked up based on a taxon name and rank. This Parser object calls it's main method, "parseTaxon", and the result is used to create a TaxonHierarchy object. This TaxonHierarchy object is used as the argument to the constructor for an RDFConverter. An RDFConverter uses the method "taxonToRDF", which takes the root element of the TaxonHierarchy and the name of an output path in which to place an RDF XML document as output. So, you can place either an absolute path or relative (to the project) path as an argument to this method. Changing RDF Namespaces The namespaces used for various predicates ("structure", "character", etc.) can be changed in the conversion/rdf.properties file. These are just populated with toy names at the moment, it's safe to change any of them (other than the "rdf" property). Changing Database Names and Table Names The filename fetching and singular->plural term processing rely on the FilenameTaxonDao and SingularPluralDao objects, respectively. These query certain databases containing tables holding relevant information. The names of the databases and tables in which the data lies can be configured in the dao/database.properties files. The properties to change to make such configurations are: - singular-plural-database - singular-plural-table-name - filename-database - filename-table-name Thus, when changes in versions occur, the table names can be easily updated.