Home | Help | Search | Login | Register

mulonga.net forum  |  Project Related  |  The Internet - Tool or threat for Tonga Culture?  |  Topic: Interview with Dr. Tunde Adegbola, African Languages Technology Initiative (Alt- « previous next »
Pages: [1] Print
Author Topic: Interview with Dr. Tunde Adegbola, African Languages Technology Initiative (Alt-  (Read 60706 times)
Peter Kuthan / AZFA
Global Moderator
Hero Member
Posts: 819

« on: February 26 2008 »

Tunde Adegbola is Executive Director of African Languages Technology Initiative, a research organisation with a mandate of making modern ICTs relevant to African languages. A person with a dual career – he is both a research scientist and a consulting engineer. Mr. Adegbola was involved in investigating the application of Cellular Automata Transforms (CAT) to psychoacoustic theory for the compression of digital audio.

His current research interests lie primarily in ICT for development and speech technologies, with particular interests in the Automatic Speech Recognition (ASR) of African tone languages. Among his achievements as a consulting engineer is the design, supply, and installation of Africa Independent Television (AIT), as well as the design of Channels Television and MITV, all in Lagos, Nigeria.

As Executive Director of Alt-i, he is actively involved in promoting multidisciplinary and inter-disciplinary cooperation between the sciences and the arts in Nigerian universities and research centers. Tunde has extensive teaching experience, having taught at the tertiary level from 1981 to the present. He taught Telecommunications for many years at the Ogun State Polytechnic (now Moshood Abiola Polytechnic) Abeokuta, Nigeria and has been teaching Artificial Intelligence and Information Networking at the post-graduate level in the Africa Regional Center for Information Science at the University of Ibadan, Nigeria since 1991.

QUeLa: Dr. Adegbola, the African Languages Technology Initiative is working both in the fields of computational linguistics and literacy improvement. How do you link these activities together?

Dr. Tunde Adegbola: As an organisation, the aim and core objectives of the African Languages Technology Initiative (Alt-i) encompass the need to make modern ICTs relevant to African Languages. As we go further into the information age, more and more human communication will be mediated by machines, and this will raise the demand, not only for humans to communicate through machines but also to communicate with machines. There is no reason whatsoever why we should be made to do this in English. In order to achieve these modes of communication in African languages however, there is a need to supplement the present objectives of the study of linguistics in African universities. Within the contexts of the linguistics of African languages, we need to develop frameworks and theories that can be passed on to and used by practitioners in Human Language Technology (HLT). To this end, Alt-i is involved in developing the relevant human and other intellectual resources to facilitate this process.

Be that as it may however, we recognise the fact that modern ICTs are almost infinitely modifiable. This is by virtue of the fact that the hardware and software components are separate and can therefore be appropriated to specific needs. For Africa and the rest of the developing world, this is a fundamental capacity that must be properly exploited. The possibility of adapting modern ICT to our specific and unique needs puts great responsibility on us to address our problems ourselves. Hence, in the process of developing the resources that are needed to make modern ICTs relevant to African languages, we also recognise, as a by-product, the capacity to address some of the debilitating problems of African societies, one of them being illiteracy. Because modern ICTs are inherently multimedia in their manipulation of information, they hold great possibilities for rejuvenating orality in Africa. We think the time for “oral literacy” has come!
QUeLa. Could you tell us more about the particular linguistic needs of Africans?

TA: Africans need to be able to live their lives in the 21st century and profit from the information age within the context of their local languages. They should not have to be able to speak English, French, or Portuguese in order to profit from the information age. There is an interesting and significant correlation between the use of a community’s mother tongue in the learning of science and technology and the capacity to achieve industrial breakthrough. We cannot ignore the manifestation of this correlation in the success of the so-called Asian Tigers. If we also observe the low state of industrialisation in African countries in which people are still struggling to learn science and technology in foreign languages, we should begin to make some salutary conclusions. Because language is the fundamental vehicle of information, the information age will surely demand a lot more from language than the industrial age did. Hence, the impediments of using foreign languages in our everyday lives is bound be more serious in the information age.

Secondly, even though one third of the languages spoken in the world are African languages, many of these languages do not have writing systems. Many of those that are written have writing systems that were not taken into consideration when the operating systems for various computer platforms were developed. Hence, apart of the traditional levels of endangerment that these languages contend with - the inability to use them on modern ICTs - imposes new levels of endangerment. The endangerment of a language also implies endangerment of the cultures they bear and by extension, the social, as well as the economic livelihoods of the people who speak them.

At yet another level, due to the multiplicity of African languages, there is usually a need for a third language (English, French, or Portuguese) before two Africans from two different linguistic realities can understand each other. This definitely will have some effect on their ability to engage synergistically.
QUeLa: In what ways do your ongoing projects address these issues?

TA: Many African languages are tone languages, and these tones are usually indicated by diacritical signs. Based on the design of most computer operating systems, it is easy to place characters next to each other but extremely difficult to place diacritical marks on top of or under characters. Some African languages even use unique letter forms that are not available in traditional computer character sets. The UNICODE consortium is working on some of these problems, but the onus still rest on the users of a language to make UNICODE work for them. These are some of the challenges of African languages that we address. We develop efficient and ergonomic computer keyboard layouts and mapping software that make it possible to type in African languages on the common desktop computer.

In addition, we believe that one of the ways that modern ICTs can benefit African languages is in machine translation. This will obviate the need for a third language when two Africans from different linguistic realities need to communicate and thereby promote and enhance mutual understanding.

Above all, we are committed to developing speech recognition and speech synthesis software for African languages. With speech recognition, an illiterate person can speak to the computer, and the computer will produce a written version of the speech. With speech synthesis software on the other hand, a computer can turn written text into speech. This holds the potentiality of redefining literacy for us.

Even though these are still largely infant technologies, we must ensure that African languages can benefit from them as soon as they mature.
QUeLa: Your idea is to reinforce peoples’ “capacity to interact with literature” – to enable them to do more than just concentrate on reading and writing abilities. What is the philosophy behind this approach?

TA: First, I must acknowledge the power of literacy. It takes information and knowledge from the aural domain to the visual domain. Certainly it does something good to the mind in that it paints pictures. However, it presents a steep learning curve to the learner, particularly to adult illiterates and thus weakens their capacity to participate in development processes. The end result is that these illiterates are excluded from the things that vitally concern them regardless of the volume of information and knowledge that they may have. The end result is that their already impoverished communities are further deprived of the wealth of their experiences just because these otherwise wise people cannot write down their experiences or read the experiences of others.

Our goal, therefore, is to include such people in development processes by providing them with alternative technologies to pen-and-paper information technology. It should be possible to provide them with computer-based systems equipped with touch-sensitive CRT, LCD, or TFT screens with attached loudspeakers, instead of the popular but exclusionary ink-stained paper. Our ambition (among many other things) is to redefine literacy, changing it from the “ability to read and write” to the “capacity to interact with literature” and thereby change the condition of African adult illiterates “from illiteracy to e-literacy”.
QUeLa: Could you give us any illustrative examples?

TA: For example, we have developed interactive information kiosks that can be used to pass important information to illiterates in their local languages. One of the most pressing information dissemination demands in Africa today relates to the need for heightened awareness of the HIV/AIDS pandemic. For many people without the ability to read and write, word of mouth - with its attendant risks of inaccurate reproduction - is about the only option. Of course radio can and has been very well used in this regard, but radio is essentially a one-way communication medium. The broadcaster usually assumes she has been heard, understood, and agreed with. With radio, useful as it is, the capacity for feedback is limited. But these interactive information kiosks not only pass vital information to their users, they also collect useful feedback information from the users for the information providers. The possibilities are almost limitless!
QUeLa: How far have you got with “Redefining Literacy”?

TA: We have obtained very useful research results in the process of laying down a strong theoretical basis for the project. We are collaborating with the Africa Regional Centre for Information Science (ARCIS) and the Department of Linguistics and African Languages, both in the University of Ibadan, Nigeria. To this end, we have been engaging post-graduate students of the University in aspects of the project and our research work is advancing, particularly in Automatic Speech Recognition (ASR) of Yoruba. In addition, one of our associates, Dr. Tunji Odejobi, of the Department of Computer Engineering at the Obafemi Awolowo University, Ile Ife, Nigeria (just about an hour’s drive away from us in Ibadan) is also working on Text-To-Speech (TTS) Synthesis of Yoruba, and we are sharing research results. With functional ASR and TTS systems, we can redefine literacy.

Even though we conceived the “Redefining Literacy” project about two years ago, we have been merely nibbling at the edges in the first two years due to lack of funding for the project. But these two years of nibbling have produced significant research results. The encouraging research results have, therefore, emboldened us to go out and seek funding for the project. Hence, we developed a five-year plan for the first phase of the project, and we have received some funding for the first year of the five-year plan from the Open Society Initiative for West Africa (OSIWA). We are now working full blast on the ASR of Yoruba and are in the process of developing a set of good practices for the ASR of other African tone languages. Based on available research results, we expect to deliver a functional speech recogniser for Yoruba by late 2007 or early 2008.

QUeLA: Dr. Adegbola, many thanks for your time.

source: http://www.elearning-africa.com/newsportal/english/news56b.php
Pages: [1] Print 
mulonga.net forum  |  Project Related  |  The Internet - Tool or threat for Tonga Culture?  |  Topic: Interview with Dr. Tunde Adegbola, African Languages Technology Initiative (Alt- « previous next »
Jump to:  

Powered by SMF 1.1.8 | SMF © 2006-2008, Simple Machines LLC