:: HanNanum Korean Morphological Analyzer & POS Tagger (Java version) ::
* Date of this document: July 31, 2010
* Last modified: July 9, 2011
* Author: Sangwon Park(hudoni@world.kaist.ac.kr)
1. Contents of the release
/JHanNanum/ Eclipse Project Directory(Source codes, Resources, Configuration files)
/GUIDemo/ GUI-based Demo application
/data.zip Resources
/conf.zip Configuration files for plug-ins
/jhannanum.jar Java archive for HanNanum
/COPYING License
/README README file
2. How to use HanNanum
- Add jhannanum.jar as a library
- Decompress data.zip at the root directory of the project (PROJECT_ROOT/data)
Decompress conf.zip at the root directory of the project (PROJECT_ROOT/conf)
(It is possible to change the location of above directories, but the path
should be specified when you use HanNanum APIs.)
- Examples:
kr.ac.kaist.swrc.jhannanum.demo.ManualWorkflowSetUp
kr.ac.kaist.swrc.jhannanum.demo.WorkflowHmmPosTagger
kr.ac.kaist.swrc.jhannanum.demo.WorkflowMorphAnalyzer
kr.ac.kaist.swrc.jhannanum.demo.WorkflowNounExtractor
kr.ac.kaist.swrc.jhannanum.demo.WorkflowSimplePos09
kr.ac.kaist.swrc.jhannanum.demo.WorkflowSimplePos22
3. GUI-based Demo
- Run GUIDemo/execute.bat or GUIDemo/execute.sh
- Environment
Java Runtime Environment 1.6 or above
4. Developing a new plug-in
- Add JHanNanum as a new Java Project in Eclipse
- Implement the Java Interface for the analysis phase and plug-in type, then the
plug-in can be used on a HanNanum work flow.
5. Update Log
:: July 9, 2011 ::
* ChartMoraphAnalyzer was updated so that an eojeol only with the beginning consonants
can be analyzed without exception.
:: June 5, 2011 ::
* JAVADOC was added. /JHanNanum/doc/
* The encoding of the all resources was changed to UTF-8.
* New plug-ins were added: SimpleMAResult09, SimpleMAResult22, SimplePOSResult09, and
SimplePOSResult22
* 5 demo programs were updated:
ManualWorkflowSetUp, WorkflowHmmPosTagger, WorkflowMorphAnalyzer, WorkflowNounExtractor,
WorkflowSimplePos22
* License (GPLv3) information is included.
:: January 10, 2011 ::
* The noun extracting plug-in after POS tagging was added. (NounExtractor)
* In the plug-in 'InformalSentenceFilter', a bug that the remaining part of the found pattern
may be omitted was fixed.
* The path for the HanNanum root directory can be set manually.
If you specify the path of the base directory at the creation of work flow, then you can
use relative path to the base directory.
Example)
> the path for the configuration files
PROJECT_HOME/data/hannanum/conf/plugin/MajorPlugin/MorphAnalyzer/ChartMorphAnalyzer.json
PROJECT_HOME/data/hannanum/conf/plugin/MajorPlugin/PosTagger/HmmPosTagger.json
> set 'data/hannanum' as the Base Directory
Workflow workflow = new Workflow("data/hannanum");
workflow.setMorphAnalyzer(new ChartMorphAnalyzer(),
"conf/plugin/MajorPlugin/MorphAnalyzer/ChartMorphAnalyzer.json");
workflow.setPosTagger(new HMMTagger(), "conf/plugin/MajorPlugin/PosTagger/HmmPosTagger.json");
:: November 16, 2010 ::
* Several tiny bugs were fixed.
* GUI-Demo application was added. It became easy to test various work flows with GUI, and people who
don't have knowledge on programing also can test them.
:: Octobor 31, 2010 ::
* New java version was released (ver.0.8 release)
* A plug-in component-based architecture was adapted.
...
