Developing Afaan Oromoo Text based Chatbot for Enhancing Maize Productivity

Abstract

Artificial neural networks are at the basis of the most state-of-the-art for a variety of activities and have enjoyed great success in a variety of tasks such as machine translation, Chatbot, image recognition, speech recognition, text summarization, and in another computation. Neural networks with recurrent model, have recently been applied to chatbot and begun to show successive results with different natural language llike English , Turkish language. How ever there is no any research conducted Afaan Oromoo text-based Chatbot. Developing Afaan Oromoo text-based Chatbot is very important, since our society have a lot problem that needs solution to solve. Since most of Ethiopia peoples are farmers there many problems that needs to solve especially in area of agricultutre.To live people must eat, to eat they must enhance their productivity , to enhance their productivity they need information how to enhance their product. In our country accessing information at right time is difficult. So, we proposed Afaan Oromoo Text-based Chatbot for enchancing maize productivity based neural network algorithm. The main goal of this research is to develop model which fit Afaan Oromoo Chatbot to solve the problem that the our society arise. We proposed two Deep learning algorithms Deep Neural Network and Recurent neural network which is based up on retrieval approach. Retrieval based Chatbot approach is approach which depend on the predefined data. We conducted human testing approach for our result, we got good performance in both model. The result from DNN model is 84.4% testing accuracy value where as experimented result from RNN based Seq2Seq model was 80% testing accuracy. Generally,the experiment result show us developing Afaan Oromoo text-based Chatbot is good solution to help farmer for enhacing their product, because farmer can communicate with the bot at any time they need.