History of machine translation in Azerbaijan

The researches carried out in the direction of machine translation in Azerbaijan can be divided into two periods; the scientific researches carried out until 2003 and from 2003 till now. During the first period (until 2003) very few research works had been dedicated to machine translation and all of them were theoretical researches. For this period as a remarkable work we can indicate “the text processing system” by M.Mahmudov. In this monograph we can come across some ideas and ready algorithms on the creating machine translation system.
      For instance: formal morphological parsing algorithm, some formal tokens of subject and predicate, methods of ambiguities and others.
      Further more we can also indicate K.Valiyeva’s, M.Mam­mado­v’s, S.Safarov’s, Z.Amirov’s research works dedicated to some aspects of problems of machine translation. They are formal syntactical parsing, formal morphological parsing and so on.
      In spite of all these researches, a machine translation system containing Azerbaijan to/from other languages was not available until recent time.
      Creating machine translation system requires more advanced researches and it is not possible to complete this work by those mentioned above, because those researches didn’t result with any model to automate machine translation system, but the main point in creating machine translation system is to develop the model of translation process.
      Any other element of machine translation system, even more important elements (for instance: automatic dictionaries, grammatical parsing algorithm, transfer algorithm and others) are secondary problems from the point of automation of machine translation system.
      From 2003, scientific works on machine translation have been started under the direction of academician A.Abbasov and the works are still continued in that field.
      The first model of machine translation system from the Azerbaijani language into other languages and vice versa has been created (A.Abbasov, A.Fatullayev). On the base of this model the first version of “Dilmanc” machine translation system was created. Also formal grammar of the Azerbaijani language (A.Fatullayev), the initial version of forming the lexical block of machine translation system was developed by A.FAtullayev, A.Mehdaliyev, F.Ahmadov and R.Fatullayev.
      In 2005-2006 ­­­­­years the research works have been continued in the frame of joined project of ministry of ICT of Azerbaijan and UNDP-Azerbaijan. (the scientific cosultant of the project is academician A. Abbasov and the project director is candidate of technical sciences A. Fatullayev)
      At different times experts M.Mahmudov, T.Abbaskuliyev, I.Tahirov, A.Baylarova, B.Talibov, R.Huseynly, I.Ashrafli, A.Galamov have also participated in this project.
      At the moment these works are continued withen the project of the Minister of Communications and Information Technologies of the Azerbaijan Republic. The first version of English-Azeri machine translation system has already copmlated and will be released at the end of 2007.
      In development of English-Azeri machine translation system are participating young specialists - R.Fatullayev, S.Mammadova, S.Shaqavatov, G.Azimova, Ch.Shirinov, E.Nuriyev and R.Zamanov.