Apertium-ara-heb

Resources

• http://www.mideastweb.org/arabic_hebrew_eng.htm tiny ara-heb-eng dictionary
• https://en.wiktionary.org/wiki/Appendix:Afro-Asiatic_Swadesh_lists whynot
• http://www.morim.com/shemi.pdf some proper nouns transliterated
• https://glosbe.com/he/ar/ translation memories and dictionaries
  • glosbe uses mostly open-source stuff … can we use what they use?
• http://web.archive.org/web/20000816051011/http://www.arava.org/Info_Center/Glossary/glossary.html glossary of environmental terms eng-heb-ara
• http://starling.rinet.ru/cgi-bin/response.cgi?root=config&morpho=0&basename=\data\semham\semet&first=1 big and scary database of semitic etymology
  • http://www.semiticroots.net/index.php?r=root/view&id=1 same thing maybe?
• http://babelnet.org/search?word=قال&lang=AR&langTrans=HE is NC
• http://compling.hss.ntu.edu.sg/omw/ – the Hebrew and Arabic wordnets are GPL and CC-BY-SA, resp.

Using http://compling.hss.ntu.edu.sg/omw/wns/arb.zip and http://compling.hss.ntu.edu.sg/omw/wns/heb.zip we can match up some thousands of entries:

$ get () { awk -vpos=$1 -F'\t' '$1~"-"pos"$" && $2=="lemma"' | sort; }
for pos in n v a r; do 
  echo -n "$pos "
  join -j1 -t$'\t' <(get $pos < arb/wn-data-arb.tab)  <(get $pos <heb/wn-data-heb.tab) | wc -l
done
n 4687
v 1316
a 115
r 82

(We may have to strip the pronounciation diacritics off the Hebrew, should be scriptable?)