Ms. Wanjiku NGANGA of the University of Nairobi believes that the latest developments in speech recognition and African languages technology will bring forth new promissing applications in the use of mobile telephones for agricultural extension services.
Active research groups on those issues need be supported:
- to improve the technological side of database queries in f.i. Kiswahili,
- to research how language technology can mediate the query (for example on diseases which affect a particular crop),
- to have database queries translated from national languages,
- to have an SQL transformation of this,
- to have the system retrieve what the farmer would want
- to translate the text-based query back into Kiswahili
- to have the query played back to the farmer in his language
Ms.Nganga gives the example of banana farmers in Kenya who use an Instant Voice Recognition (IVR) system to ask his question. The system plays back the response in Kiswahili to the farmer.
She recognizes that language technology is very challenging and that a lot of work still needs to be done. Many present applications are SMS based. The language aspect is bringing things a step further: "We can reach farmers regardless of their local expertise in different languages"
She finally elaborates on the present collaboration between the MIT and the UNBI on training for mobile phone programming and how to develop a number of new applications for the mobile phone.