Logo Lanfrica

tasuqilt/LangDetect_Kab_Files

Domain:

natural language processing

Record type:

software
Creator:
tas
Host:
Adding support for Kabyle language (ISO 639-2 code “kab”) # LangDetect_Kab_Files Adding support for Kabyle language (ISO 639-2 code “kab”) I have added the language profile "kab" (obtained with the command java -jar langdetect.jar --genprofile) Profile created using a Wikipedia export from kabwiki-20220120-abstract.xml.gz from wikipedia dumps site: dumps.wikimedia.org Installation (how to use "kab" in language detection): 1. Install "LangDetect" using github.com 2. Add profile "Kab" file store here in this repository to the "LangDetect" installation folder 3 Testing: Run command "Python detect_lang.py" Details for generating "kab" language profile (you only need to do this in case you want te regenerate another "kab" profile) Steps provided here: code.google.com Downloaded langdetect.jar Download and use new jsonic version (>1.2.5) from java2s.com (to fix error Exception in thread "main" java.lang.NoClassDefFoundError: net/arnx/jsonic/JSONException) Run command java -jar lib/langdetect.jar --genprofile -d [directory path] [language codes] Make sure that [directory path] contains only the file kabwiki-20220120-abstract.xml.gz and nothing else. Under [directory path], you need to have a subfolder named “profiles” that will store the generated language profile “kab” Misc Troubeshooting: github.com Note: In linux, you need to copy the profile generated “kab” into installation folder of LangDetect folder such as : .../lib/python3.7/site-packages/langdetect/profiles