This is the top level page for a VKC documentation project on LexEVS loaders. This page and all child pages will move to the LexEVS space once it is created.
LexEVS provides both a GUI and a command line interface to load terminologies. While the GUI is very functional, many system administrators prefer the command line interface so that scripts can be adjusted to increase memory and tune other java virtual machine settings to insure loads of larger terminologies load with adequate resources. Typically a user may select a loading script, open it in an editor, and increase java heap size and PermGen memory, depending on the machine’s resources, and save the script before running with the appropriate options written into the command line. Still, the GUI can be convenient for loading smaller terminologies and in many cases works fine for loading moderately large terminologies like the NCI Thesaurus. Loading resources requires some knowledge of the source.
Most source loads can be easily accomplished by pointing either the command line script or the lbGUI interface at the source file and running the loader. Generic loading instructions can be found in the administrative gui guide or the administrative command line guide, but for many sources you can use a variation of the following command line script:
Linux
./LoadOWL.sh -in "file:///ontologies/owl/amino-acid.owl"
Windows
LoadOWL.bat -in "file:///ontologies/owl/amino-acid.owl"
Substituting the loader script for the needed format and any path needed to point the loader to a local source file.
Some sources are special cases and need special handling. Included in this category are the NCI Thesaurus in OWL format and any files loaded from UMLS RRF formatted sources. The NCI MetaThesaurus is the largest terminology we load and as such it also requires special handling. OWL terminologies do not normally require special handling, but LexEVS offers some advanced loading options users can take advantage of. At the bottom of the page we link to tutorials for each of these.
Loading larger terminologies can be very time and resource consuming and this can be helped by following recommendations for database optimization and for proper configuration of the lbconfig.props file.
The default setting for the value for the database primary key is the following:
Because this is very taxing on the index processing at the end of the load, we recommend changing it to SEQUENTIAL_INTEGER unless you have a priority need for Global Unique Identifiers.
The guide includes the following sections:
