AgrotIJet e-Advisor for.Small, Scale Farmers
Abstract
A lot has been done in using Information and Communication Technology (K'T) in business yet
very little in solving agricultural problems in developing countries. Small scale farmers in
developing countries lack adequate knowledge to interpret and understand weather information.
Many weather dependent variables like rainfall, humidity, and temperature are not clearly
understood by farmers. Farmers lack adequate knowledge to make use of weather variables in
managing their crops. As a result, small scale farmers make decisions on grounds of uncertainties
when carrying out operations like scheduling planting time, chemical application, crop nutrition
and weeding making crop yields to be lower than expected.
This project tried to find the most suitable way of processing the available data and
disseminating agro-meteorological information to small scale farmers. In order to understand the
underlying issues a survey was conducted to collect data from small-scale farmers,
meteorologists, agricultural extension officers and a research station.
Different knowledge management approaches were surveyed on their usability for processing
and disseminating agro-meteorological information to small scale farmers. An Expert System
was designed and its knowledge base populated with knowledge and intelligence collected from
the field. The user interface design uses internet and mobile phone technology.
After implementing the prototype, some evaluations were made thorough companson with
human expert. The aim was to study whether the approach used really helped to solve knowledge
management problem for small scale farmers. In these tests it emerged that the concept of
knowledge management is a useful approach of processing the available data and disseminating
agro-meteorological information to small scale farmers. "Through evaluation tests conducted on
the prototype, it was showed that a system developed out of the designed model would enjoy
better accuracy levels of up to 66% in crop suitability prediction, 75% advance warning and 72%
yield prediction.
Citation
Master of sciencePublisher
University of Nairobi School of Computing and Informatics, University of Nairobi