Quranic Arabic Corpus

The Quranic Arabic Corpus is an annotated online linguistic resource covering the 77,430 words of the Qur'an, meant to give researchers morphological and syntactic annotation of the language of the Qur'an. ¹

Each word is tagged with its part of speech and several morphological features, such as whether it is a noun or a verb, masculine or feminine. Unlike other annotated Arabic corpora, it adopts the framework of traditional Arabic grammar, i'rab, and shows it as dependency graphs. ¹ It includes a manually verified part-of-speech tagged corpus, a treebank, morphological search, a machine-readable morphological lexicon of Qur'anic words into English, a part-of-speech concordance by lemma and a message board for volunteer annotation. ¹ It gives a word-by-word translation drawn from accepted English sources rather than a new translation. ¹

The work began with automatic part-of-speech tagging by Arabic language technology; each of the 77,430 words was then reviewed in stages by two annotators, and improvement continues. ¹ The project is led by Kais Dukes at the University of Leeds, in the Arabic language computing group of the School of Computing, supervised by Eric Atwell; its first release came in November 2009. ¹

The corpus has been used to train Arabic part-of-speech taggers, to categorise the suras of the Qur'an automatically and to analyse the text's pauses and prosody. ¹

The site is at corpus.quran.com and has articles in the Arabic, English, Indonesian, Malay and Russian Wikipedias. ² The museum quotes none of the Qur'anic text here.

• English Wikipedia, "Quranic Arabic Corpus", revision 1375079606. ¹
• Wikidata Q7272925. ²