Intelligent Recommendation System Based on Image Processing

Abstract

Recommendations for images are the core tasks of many website recommendation systems. This paper develops a relatively complete image recommendation system based on the relative data analysis of the hotel website. In this paper, the image processing methods such as computer vision in the field of deep learning are applied to the information processing classification of the image of the hotel website. The image recommendation system established in this paper is more targeted than the recommendation system of some previous websites. This paper evaluates the aesthetics of NIMA images of the photos of the houses, and uses the show and tell model to describe the image and text, and finally according to the star rating and user preferences. Use collaborative filtering algorithms for house listings and recommendations. The accuracy of the recommendation is greatly improved, and the quality of the service selected by the user can be effectively guaranteed.