OWLIM 3.5 released - delivering improved performance, improved functionality and online documentation

[Apologies for cross-posting]

Ontotext are pleased to announce version 3.5 of their OWLIM 
<http://www.ontotext.com/owlim/index.html> family of semantic 
repositories. This release further improves Lucene-based full-text 
search, provides remote-notifications for both BigOWLIM and Replication 
Cluster and makes all BigOWLIM advanced features available through the 
Jena RDF framework <http://jena.sourceforge.net/>.

Recent improvements in BigOWLIM, e.g. optimisations for queries that 
combine ORDER BY with LIMIT/OFFSET, allowed it to demonstrate 
outstanding performance on the latest independent evaluation performed 
by the authors of BSBM 
<http://www4.wiwiss.fu-berlin.de/bizer/BerlinSPARQLBenchmark/spec/index.html>. 
A beta-version of BigOWLIM 3.5 was used for the evaluation conducted in 
February 
<http://www4.wiwiss.fu-berlin.de/bizer/BerlinSPARQLBenchmark/results/V6/index.html> 
in which BigOWLIM demonstrated by far the best loading performance and 
also the best query performance among those repositories that can handle 
update and multi-client query tasks.

Another important change that applies to all editions of OWLIM is that 
the latest documentation is now provided online 
<http://owlim.ontotext.com/>. PDF versions of user guides will still be 
provided in the distribution zip files as a courtesy, but the main 
reference point is now the website.

Many other updates and fixes are also included. Further details follow:


  BigOWLIM

    * *Remote notifications:* A new mechanism to complement the existing
      high-performance 'in-process' notification mechanism. This new
      mechanism allows clients to subscribe for the given statement
      patterns to remote BigOWLIM repository instances.
    * *Schema editing:* Read-only schemas loaded at database
      initialisation time allow very fast deletion of (instance)
      statements by using the 'fact-retraction' method that computes the
      necessary inferred statements to delete. A new mechanism is
      provided with this release that allows 'read-only' schema
      statements to be modified when necessary.
    * *Query optimisations:* Several improvements have been made to
      query optimisation, including the special case when using ORDER BY
      with LIMIT/OFFSET.
    * *Configuration spreadsheet tool:* The memory calculator from
      previous versions has been updated to estimate appropriate
      BigOWLIM configurations for the specified hardware, dataset
      characteristics and selected features.
    * *Storage files updated automatically:* There are minor differences
      in storage file formats between versions. Versions of files back
      to 3.1 are now detected and updated automatically.
    * *owl:sameAs optimisation can be disabled:* The owl:sameAs
      optimisation can now be switched off using the disable-sameAs
      configuration parameter. This update might be useful when using
      the empty or rdfs rulesets.
    * *Lucene-base full-text search enhancements:* Even more
      fine-grained control over what to include in the indexed RDF
      molecule. Separate include/exclude lists are now supported for
      both predicates traversed and entities visited.
    * *All OWLIM plug-ins available with Jena interface:* All the
      BigOWLIM advanced features are now fully supported when using
      BigOWLIM with the Jena framework. This includes RDF Rank, RDF
      Search, Node Search, RDF Priming and Geo-spatial extensions.


  OWLIM Replication Cluster

    * *Write-only worker node:* When worker nodes are added to the
      cluster via the JMX interface, they can be specified as being
      'write-only'. These nodes will be kept in synch with the rest of
      the cluster, but will not take part in answering cluster queries.
      The motivation for this feature is to have one or more worker
      nodes available for batch processing of queries that do not affect
      the overall query performance of the cluster.
    * *Remote notifications:* A new mechanism to complement the existing
      high-performance 'in-process' notification mechanism. This new
      mechanism allows clients to subscribe for the given statement
      patterns to OWLIM Replication Cluster master nodes.


  SwiftOWLIM

    * *Artificial limit on ruleset size removed:* Very large custom
      rule-sets were causing out of memory exceptions or problems during
      compilation. The size of rule-sets is now practically unlimited.
    * *Bug Fix - Data loss from abnormal termination:* This fix prevents
      data from being lost after two successive abnormal terminations.
      This was due to a misidentification of backup files.
    * *Bug Fix - Synchronisation problem:* A synchronisation problem led
      to intermittent incorrect query answering of the LUBM-1 functional
      test. The issue appeared at an average of 1 of 80 runs and it was
      due to bad read-write synchronisation when adding to collections.

The OWLIM Team, March 2011