Chatbot Using Deep Learning
Academic Leadership - Online Journal,
Vol. 21 No. 6 (2020): Vol-21-Issue-6-June-2020
Intelligent Chatbot is a system which can interact with humans and answers questions on a certain domain. Today, the challenge is to build a system which will resemble human brain. Generally, the brain stores the memory in a decentralized manner across the brain with the help of neuron as opposed to a centralized manner in computer file system. There are short term and long-term memory storage with different priority based on variety of situation. The system can take inputs in written or voice format and respond the question from a knowledge base. In most cases a chat bot does not have problem solving capabilities. Our system can solve data structure problems using Deep learning mechanism. Deep neural network (DNN). With a given dataset the system can provide services to access data in format such as arrays, stacks, queues and trees. Based on these data structures I can solve problems like traversing lists, reversing numbers and translating language of syntactic divergences. The learning service is not an algorithmic program rather a trained model using DNN. With the implementation of problem solving chatbot, it will understand how to organize and retrieve data based on user’s data structure choice. I used Neural Stack Machine (NSM) with Recurrent Neural Network (RNN) as the controller.