Skip to end of metadata
Go to start of metadata

This section discusses how to configure the performance of the semantic repository.

If you have any custom ontology extensions, place them in the %KIM_CONTEXT%/kb/owl folder. For more information, see the detailed guide about extending the ontology.

Configure <KIM_HOME>/config/owlim.ttl in the following way:

The most important parameters given above are:

  • entity-index-size
    • Definition (OWLIM SE documentation): Defines the number of entity hash table index entries.
      The bigger the size, the less the collisions in the hash table, and the faster the entity retrieval. The entity hash table does not rehash, so its index size is constant throughout the repository life.
    • How to set it in the context of KIM: The entity-index-size should be equal or a little larger then the maximum number of expected RDF entities.

The number can be approximated as follows:
RDF entities = Named Entities in pre-loaded KB * 4 + Documents * 8
Since the entity-index-size cannot be changed once it is set, you should expect maximum number of documents to be indexed in the server. However, this is not a hard limit - there will be only a decrease in performance, if the number of RDF entities exceeds the entity-index-size, and the penalty will be proportional to the difference.

  • tuple-index-memory
    • Definition (OWLIM SE documentation): The amount of memory to be distributed among RDF statement cache.
    • How to set it in the context of KIM: When using the server only for populating documents, we recommend setting the parameter to a smaller value (~50 MB), in order to allocate more memory for semantic annotation. On a static server, used only for querying, you can disable the semantic annotation and try increasing that cache to improve performance.
  • predicate-memory
    • Definition (OWLIM SE documentation): Memory used for predicate list cache.
    • How to set it in the context of KIM: Keep the default value. This cache only affect a small fraction of the queries used in the KIM interface.
  • fts-memory
    • Definition (OWLIM SE documentation): Memory used for full-text index cache
    • How to set it in the context of KIM: The advice about tuple-index-memory applies to this cache as well.

For details on configuration options, see the OWLIM documentation.

Labels:
conf_repos37 conf_repos37 Delete
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.