ZenDen - A Personalized House Searching Application
2020 IEEE Sixth International Conference on Big Data Computing Service and Applications (BigDataService)
A method directly linking buyers and sellers in the housing market would benefit both parties in both the renting and purchasing domains. The real estate mobile applications available today, such as Zillow and Redfin, are highly dependent on filter based searches (location, price, bedrooms, etc) and contain many screens that require zooming in and out of map-based interfaces. Furthermore, the student market and lower-income demographics that are more likely renting a house instead of buying are usually not the focus of these applications. Described in this paper is a novel mobile application that will change the way people in the renting market, e.g., university students, find lodging. The implementation contains a streamlined swipe-based user interface backed by a user-house recommender system to manage content. Deep learning techniques are used in building the recommender system that recommends houses to users based on their view history. For image classification, we build convolutional neural networks (CNN) for analyzing house images. The goal is to create a personalized, easy to use application that will reduce the effort and time required for people in the renting market to find housing.
CNN, Microservice Architecture, Mobile Housing App, Recommender Systems
Kristina Milkovich, Saurabh Shirur, Pratap Kishore Desai, Likhith Manjunath, and Wencen Wu. "ZenDen - A Personalized House Searching Application" 2020 IEEE Sixth International Conference on Big Data Computing Service and Applications (BigDataService) (2020): 173-178. https://doi.org/10.1109/BigDataService49289.2020.00034