Photo by thinkmagically from unsplash
A conversational agent (chatbot) is a software that can communicate with humans using natural language. Conversation modeling is an extremely important topic in natural language processing and artificial intelligence (AI).… Click to show full abstract
A conversational agent (chatbot) is a software that can communicate with humans using natural language. Conversation modeling is an extremely important topic in natural language processing and artificial intelligence (AI). Indeed, since the birth of AI, creating a good chatbot remains one of the most difficult challenges in this field. Although chatbots can be used for a variety of tasks, they generally need to understand what users are saying and to provide appropriate answers to their questions. In this paper, we present midoBot: a deep learning Arabic chatbot based on the seq2seq model. midoBot is capable of conversing with humans on popular conversation topics through text. We built the model and tested it in the Tensorflow 2 deep learning framework using the most seq 2 seq Model architectures. We use a dataset of ~81,659 pairs of conversations created manually and without any handcrafted rules. Our algorithm was trained on a VM on google cloud (GPU TESLA K80 10 GO). The results obtained are significant, In most questions the chatbot was able to reproduce good answers.