Deep Learning based Chatbot Utility to assist Employees in identifying the Safety Risks based on the Classification of the Industry Incidents

The advancement in technology has improved the interaction between people and devices in many ways. One such way is a chatbot. Chatbot is an artificial intelligence-based computer program that uses messaging platforms to interact with people. With the evolution of technology, the manpower in industries has been reduced considerably. However, it is still a long way before industries achieve 100% automation. In industries, accidents and injuries are common despite all the safety measures and precautions put in place. It is imperative to classify industrial incidents into distinct categories to ascertain whether the incidence was purely an accident, the result of carelessness or incompetence, or if the causes can be traced back to the industry’s procedures or products. This minimizes the effects, prevents recurrence, and lowers the frequency and intensity of incidents. To achieve this, we employ exploratory data analysis (EDA) on a dataset from major Brazilian industries to determine the most common causes of accidents, types of accidents, timestamps, types of employee injuries, and other factors. This paper aims to develop a chatbot utility that is ML/DL-based and can assist professionals in identifying safety risks.