You can download the code or the compiled application at the JoBimText Sourceforge project page. The most recent version of the pipeline is 0.1.2. It can be downloaded directly using this link.
For older pipelines, check the files section on Sourceforge.
Some components are distributed under the GPL license. They can be found in a separate JoBimText-GPL project. The most recent JoBimText pipeline with GPL components can be downloaded from this location: jobimtext_pipeline_0.1.2.tar.gz
If you do not have access to a Apache Hadoop cluster or want to use existing Distributional Thesauri, you can download and use precomputed models. Most of the models are listed on the models page at Sourceforge. Models computed from Google Syntactic Ngrams are available on their separate page.
The compiled .jar files are available for download on Sourceforge. According to your needs, you can either take the ASL-licensed org.jobimtext.evaluation.jar or the GPL-licensed org.jobimtext.evaluation-gpl.jar.
The difference between these two files is that only the GPL version contains software to calculate similarity scores. Therefore, we recommend to use this version, if you are free to use GPL software.
The compiled .jar file is available for download: ChiWhiDisamb.jar
You can view the options by opening it with a -h parameter. Consult the introduction to sense clustering to see its usage.