Abstract: In this paper, we have proposed a recipe recommendation system that makes use of images of ingredients to recommend a recipe to a person who doesn’t know anything about the contents and its proportion in a particular recipe. This system is also beneficial for health-conscious people. With the changing living manners, diet habits are changed and work load has increased which resulted in various diseases, such as diabetes, BP, problems related to heart and so on. All these diseases can be controlled by avoiding uneven and unhealthy food. So it’s important to understand that what is the proper diet and in how much quantity it should be taken. Our main aim is to recommend recipes to maintain their health for people with disease and without disease which will satisfy the needs of user. To give the recommendations, recommended system uses the user’s profile, their favourite food and  details of that food. Clients from various countries, belonging to various cultures are contributing in terms of a large number of new recipes on the web all over the world.  Every recipe contains of so many distinct elements. Thus, user might not be able to identify all the ingredients or contents. In this paper we propose a recommendation system for recipe using Convolutional Neural Network (CNN) which is used for supervised learning in order to analyse the data. Recommended Recipe contains diverse ingredients, cooking procedure, categories and so on. The recipe which includes the ingredients mentioned by user with proper nutritional values will be a good recommendation.

Keywords: Image processing, Machine learning, Recommendation system, Collaborative Filtering, Ingredients, CNN

PDF | DOI: 10.17148/IJARCCE.2019.8308

Open chat
Chat with IJARCCE