Arabic natural language processing: handwriting recognition - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2008

Arabic natural language processing: handwriting recognition

Abdel Belaïd
  • Function : Author


The automatic recognition of Arabic writing is a very young research discipline with very challenging and significant problems. Indeed, with the air of the Internet, of Multimedia, the recognition of Arabic is useful to contributing like its close disciplines, Latin writing recognition, speech recognition and Vision processing, in current applications around digital libraries, document security and in numerical data processing in general. Arabic is a Semitic language spoken and understood in various forms by millions of people throughout the Middle East and in Africa, and it is used by 234 million people worldwide. Furthermore, Arabic gave rise to several other alphabets like Farsi or Urdu increasing much the interest of this script. Farsi is the main language used in Iran and Afghanistan, and it is spoken by more than 110 million people, concerning also some people in Tajikistan, and Pakistan. Urdu is an Indo-Aryan language with about 104 million speakers. It is the national language of Pakistan and is closely related to Hindi, though a lot of Urdu vocabulary comes from Persian and Arabic, which is not the case for Hindi. Urdu has been written with a version of the Perso-Arabic script since the 12th century and is normally written in Nastaliq style.
Fichier principal
Vignette du fichier
belaid-acit-keynote.pdf (91.44 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

inria-00347229 , version 1 (15-12-2008)


  • HAL Id : inria-00347229 , version 1


Abdel Belaïd. Arabic natural language processing: handwriting recognition. International Arab Conference on Information Technology - ACIT'2008, Dec 2008, Hammamet, Tunisia. ⟨inria-00347229⟩
121 View
198 Download


Gmail Facebook Twitter LinkedIn More