Recipe Bot: The Application of Conversational AI in Home Cooking Assistant

Conversational Artificial Intelligence (AI), which allows people to have human-like interactions with computers, has experienced a boom in recent years. A wide range of fields such as healthcare, finance, and retail have applied conversational AI in their websites to save efforts of completing easy tasks and provide voice interaction with end-users. It enables customers to find quick answers to frequently asked questions and service providers to save time to tackle more complex problems. This paper introduces the Recipe Bot, which is a conversational agent that provides matching recipes based on the information given by users. The intention of Recipe Bot is to help users get rid of unused ingredients in the fridge by providing related recipes. Recipe Bot allows users to input a specific dish name or provide a region, type, and/or ingredients of the food they would like to have, and then it returns a recipe list according to sorting and nutrients requirements given by users. The chatbot is built using Google Dialogflow platform to recognize the user’s intentions and Spoonacular API to find recipes that match the search query. This paper discusses Recipe Bot’s architecture, functionality, and inadequacies to be improved. It gives a detailed example of the interaction between the chatbot and a user, which demonstrates how the user interface will simplify the problem of finding a user-interested recipe.