Study On Data Warehousing Techniques For Real-Estate Price Analysis

The historical process of social and economical development of human being makes it clear that the real-estate industry is the forerunner and underpinning to the national economy. With the acceleration of nation's industrialization and urbanization, the significance of real estate has been more and more remarkable. Investment in real estate is an integrated part of investment in fixed assets and a primary driving force to the development of national economy making great contribution to GDP. Since the year 2002, the development of real estate, which has high degree of correlation and strong driving power to other sectors, has grown generally in a reasonable and healthy fashion. However, due to overly fast investment growth, land supply and price hike in some places of the country, problems with overheating and some other structural problems are emerging posing huge risk on the market. To scientifically and objectively evaluate and predict the trend of real estate market making investors avoid of market risks and keep real-estate market grows healthily and stably holds great theoretical and practical significance.A warehouse integrates business data from different nodes in enterprise's network. Before separating business operational environment from information analysis environment, it collects, inducts and processes data supporting decision making. In doing this, warehouse provides decision-makers with groundwork. OLAP is a widely applied tool for analyzing information in warehouse. With OLAP, users can seek data stored in warehouse from a multidimensional view in a quick, consistent and interact way. Meanwhile, the development of data mining techniques deepens applications of warehouse. Users using data mining models can find out patterns and rules, get insights into data, master hidden laws, learn knowledge and predict future events quite precisely.This thesis overviews the fundamentals of warehousing and data mining techniques, designs a simple index system of real estate price and a composite index for predication and applies warehousing techniques to the analysis and prediction of the composite index. Then, on the platform of SQL Server 2005, the warehouse of real estate price is built. The ETL is practiced with Integration Services. Thus, the multidimensional data set is constructed on the basis of real estate warehouse. Beyond that, future price prediction and its impact factors analysis are made through Microsoft Time Series and Artificial Neural Network. Finally, a report on result of OLAP is derived.