Natural language processing for real - estate application

Abstract

Chatbot development is popular due to its benefits for e-commerce applications. Especially, real-estate applications have shown the need of artificial assistant services through its helps on simple tasks (e.g. Questions and Answers). Chatbot is a product of machine learning which is implemented by applying the natural language processing methods. However, it is difficult for the bot to understand completely the conversation and do next actions just by matching general patterns. Many machine learning appoachs are introduced to solve the problems. Therefore, in this thesis, natural language processing (NLP) models, which are strongly recommended for chatbot development, will be experienced and evaluated. This approach helps developer to understand the bot production processes and have brief knowledge of such models. However, instead of implementing all the machine learning model in detail by code, a high-level framework will be used to ease the development process and replace duplicated implementations. Within this frame work, all the related architecture of general chatbot are shown. Finally, bot will be integrated into one of social platforms (i.e. Facebook, Slack, etc.) for free trials.