Research On The Classification Of Second-hand Housing Resources In Beijing From The Perspective Of The Buyers' Preference

As the capital of China,Beijing has many superior livelihood resources such as high-quality education,advanced medical care and so on,attracting more and more people to stay in.The price of Beijing's first-hand housing stays high,and the second-hand housing market is more active than the first-hand housing market on account of the driving rigid demand.After the prosperity in 2015 and 2016,the recession in 2017,the transaction of the second-hand housing transaction market upsurged in the first half of 2018,while the volume and price both fell in the second half of the year.It can be found that the market shows great fluctuations due to the policy regulation,which has a negative impact on buyers' s willingness to make a purchase.This dissertation adopts ordered logistic regression and the machine learning method to effectively classify second-hand housing in Beijing based on buyers' preferences,and to put forward corresponding countermeasures for the marketing and development of second-hand housing trading platform in order to promote the benign and sustainable development of the second-hand housing transaction market.First of all,the dissertation analyzed the current development status of the second-hand housing transaction market in Beijing,and compared the performance of the market from the 13 urban districts.In the second place,taking the LianJia platform as an example,the paper applied the text mining to the commercially aviliable second-hand housing by using web crawler technology and obtained the main influence factors to classify the house resources after analyzing.The dissertation used transaction cycle and the number of followers as indicators to measure user's preferences to classify and mark the second-hand houses into three categories: hot housing,ordinary housing,and unsalable housing.The relationship between the category of houses and the attribute variables of houses was explored through the contingency table and correlation coefficient analysis.Furthermore,it used the ordered triclassification logistic regression,random forest and GBDT model to fit the housing datasets to obtain coefficient,odds ratio,feature importance,and analyzed the impact of each influence factor on the housing level and then estimated the classification result.Through the study mentioned above,we can draw a conclusion that the house would be sold well if it has a low price,a small area,the less bedroom and living room,convenient elevators,biggish ladder ratio,surrounding subway,mixed building structure,board building and simple decoration,and it's better if there are more than 5 years of house ownership.On the contrary,the disadvatages of selling houses are the high price,the large area and the exquisite decoration.In addition,the five urban areas have their own characteristics: Chaoyang District and neighboring Fengtai District have similar user preferences.The housing with the high total floor,located on top and low floor are more likely to sell well.Haidian District is similar to the adjacent Changping District.The top and middle floors will enhance the popularity of the housing.The difference is that the lower the total floor of Haidian District,the more popular,but the higher the total floor of Changping District,the more popular.The more living rooms,kitchens and washrooms in Changping District,the easier it is to sell.On the one hand,as one of the largest residential areas in China,the housing price of Changping is relatively appropriate for the civilian.On the other hand,many of the houses are rent to outsiders who do not have a house in the local district.The house would be easier to rent to multiple person and increase the rental income if there are muli-kitchens in the housing.Xicheng District is quite special.The popularity of brick and wood structure is about 1.10 times that of houses without the structure of brick and wood.The longer time the housing is,the more popular it is.In view of the housing with old Beijing characteristics in the East District and West Districts,some buyers may buy a house because of the old Beijing traditional culture.It's may be that Xicheng District has small population and high housing prices,so the the house with the small ladder ratio than those have a big ladder ratio.The research conclusions mentiond above provide more detailed house-purchase needs for the marketing personnel.And they have definitely realistic significance for the development of the second-hand housing transaction market.