【作者】 李明铭；

【导师】 沈陈华；

【作者基本信息】 南京师范大学 ， 地图学与地理信息系统， 2011， 硕士

【摘要】 近年来,随着地方财政对土地出让依赖的增强,房地产开发商对房地产行业暴利的追逐,广大消费者对房地产市场交易的渴望和房地产投资者对房地产市场交易的热衷,国内大中城市房地产市场日益繁荣。房价成为每天的焦点话题,而国家发布的房产价格指数又总与大众感受产生偏差。社会公众需要专业、权威的房产交易信息网络发布平台,公开提供实时的房产交易信息；根据实时的房产交易信息,编制更符合大众感受的房产价格指数,分析房地产市场的价格走向；利用网络空间服务,辅助用户从空间区位角度,更直观、精确地了解房地产市场的宏观发展与趋势。本研究在充分分析我国现有房产价格指数编制和现有房产交易信息网络发布平台存在的不足之后,综合利用网络爬虫技术、网页解析技术、Web技术和GIS技术,进行房产交易信息的网络提取、网络空间发布与空间分析,并基于Hedonic模型进行了南京市六城区二手住宅价格指数的编制,具体的研究内容和成果主要包括：(1)基于Hedonic模型的房产价格指数编制基于房产独特的空间特性,对现有Hedonic模型进行优化,扩展模型变量,调整量化方式,确定模型的函数形式；变更房产价格指数编制所需的属性数据和空间数据的来源；基于实验样本对Hedonic模型进行估算和模型修正。(2)房产网络交易数据的收集、提取与量化利用网络爬虫程序收集房产交易网页；归纳网页布局规则并基于HTMLParser的网页解析技术,提取房产价格指数编制所需的属性数据；对提取的房产网络交易数据进行质量检核和量化。(3)房产交易信息的网络空间定位、发布和分析基于网络空间服务构建房产交易信息网络发布平台,在平台空间定位并发布网络收集与提取的实时房产交易数据；通过平台拓展的空间分析功能,获取房产价格指数编制所需的空间数据,空间可视化表达房产空间分布与价格分布,辅助公众决策。(4)南京市六城区二手住宅价格指数编制实例利用上述获取的属性数据和空间数据,基于Hedonic模型对南京市六城区的二手住宅,进行分区房产价格指数的编制；通过与“七十个大中城市二手住宅分类价格指数”的比较,本研究的编制结果没有严重脱离权威指数,且更符合大众感受。更多还原

【Abstract】 In recent years, with the local Government’s financial dependence on the land, Real estate developers’ pursuit of profits on the real estate industry and the majority of real estate consumers’ interestes in the property market transactions, the domestic real estate market is booming. House price is becoming the daily hot topic. House price index issued by the State is always inconsistent with the public feeling.The public need for a professional, authoritative web publishing platform to publish available real-time information on real estate transactions. According to the information, people can calculate the house price index which is more line with the public feeling to analyze the real estate market trends. Web map service which is offered by the platform will help users to understand the macro development and the trends of the real estate market directly and precisely from the spatial view.After full analysis of the shortcomings of current real estate price index and the existing web publishing platform on real estate transaction information, this study combined the technologies of web crawler, page parser, Web and GIS to calculate the price index of second-hand residential property in Nanjing city based on the Hedonic model.Specific research contents and results include:(1) House price index construction based on Hedonic modelFor the unique spatial characteristics of real estate, this study extended model variables, adjusted the quantitative methods and chose the model’s functional form to optimize the existing Hedonic model. It changed data sources of the attribute data and spatial data which are required for the house price index construction. It estimated the Hedonic model and updated it based on the experimental samples.(2) Real estate web transaction data collection and extractionThe study collected web pages of real estate transactions using the web crawler, summarized the layout rules of the pages, extracted the attribute data which is required for the house price index construction using HTMLParser, and checked the quality of the extracted data and quantified them.(3) Construction of web publishing platform on real estate transactionsThe study constrcted a web publishing platform based on web map service to publish the extracted real estate transaction data in real time. The study used the expanded spatial analysis functions of the platform to obtain and verify the spatial data which is required for the house price index construction. At the same time, the platform also provided the function to spatial visualize and distributed the house price to support decision-making.(4) Price index construction examples of Nanjing cityThe study calculated the price index of second-hand residential property in six districts of Nanjing city based on the Hedonic model using the data obtained above. Compared with the "Classified Second-Hand Housing Price Index of Seventy Cities" published by National Bureau of Statistics of China, the results of this study is not out of the authority but more line with the feeling of the general public.更多还原