Indonesian POS Tagger

INDRA makes use of the Indonesian POS Tagger (Rashel et al., 2014), a tool developed by a team from the Faculty of Computer Science of the University of Indonesia, for POS tagging in Indonesian. This tool is not included in the INDRA repository, and users are responsible for downloading, installing, and setting up the tool.

Indonesian POS Tagger is based on the Indonesian Morphological Analyzer (MorphInd) (Larasati et al., 2011), a tool which handles both morphological analysis and lemmatization for a given surface word form.


0. Unix Operating System

1. Install Perl

2. If your machine does not have java, install java.

$ sudo apt-get install default-jre

3. Download foma

Please make sure foma was registered in your PATH environment variable.

4. Download Indonesian Morphological Analyzer (MorphInd)

Please make sure to put it in morphind folder.


The version of java and javac have to be the same in order to run the Indonesian POS Tagger.

$ java -version
java version "1.7.0_80"

$ javac -version
javac 1.7.0_80

If not, you have to svn up the logon.

$ cd ~/logon
$ svn up
$ rm -f bin/java
$ rm -f bin/chasen
$ rm -f naist/bin/linux.x86.32/chasen
$ svn cleanup *

If it does not work, delete the logon tree and reinstall. Note that before deleting the logon tree, your own data such as tsdb have to be stored.

$ rm -rf logon
$ svn co logon

This takes a couple of hours, then remove the java in logon.

$rm -f logon/bin/java

Reset up the FFTB for INDRA (see IndraTreebanking).

IndraPreprocessing (last edited 2015-07-06 12:48:28 by DavidMoeljadi)

(The DELPH-IN infrastructure is hosted at the University of Oslo)