Data-Intensive Real-Estate Information Processing and Display through an Integrated Online Service

点击翻译按钮获取中文摘要

1527058 Alberta Inc. is developing an online service that processes, analyzes, summarizes and displays real-estate property pricing data and related information to end users. The system will mainly consist of two parts: a data-intensive backend system and a novel graphic user interface for information display in the frontend. As an essential part of the information processing and visualization application, the backend system will take public real-estate property assessment information and recent real-estate transaction information on MLS listings as inputs, compute its own price estimate for each property using statistical learning methods. Our research group will design effective real-estate property price estimation algorithms based on a novel application of latent factor models and sparse matrix completion, which are among the most recent advancements in statistical learning theory. We will also design the statistical procedures to provide aggregate summary of the market trend on multiple levels pertaining to the communities, neighbourhoods and submarkets. Then, we will collaborate with the partner company to incorporate the developed algorithms into an online software system that automatically retrieves, analyses and displays data. A novel graphic user interface will be designed and developed in the forms of both a web service and a mobile app to facilitate data visualization on micro as well as macro scales.