基于网络搜索数据的湖北省社会消费品零售总额的预测研究
|
摘要
本文基于湖北省的数据,研究网络搜索数据能否优化以往仅使用政府统计指标对社会消费品零售总额的预测。为此,建立6种模型,分别以政府统计数据、网络搜索数据或它们的不同组合做解释变量。实证结果显示,网络搜索数据的加入显著地改进了仅用政府统计指标做解释变量模型的预测效果,网络搜索数据是政府统计数据的良好补充,需要注意的是,须将其和政府统计数据加以区分使用。本文的研究结果可助力湖北省社会消费品零售总额的统计工作,也能为政府部门制定拉动内需政策提供一定参考。
In the era of big data, web search data is generated at an alarming rate every day, which enriches the type and source of data. There is still little literature involved about whether and how to use web search data to predict the total retail sales of social consumer goods. Based on the data of Hubei Province, this paper researches whether the web search data can optimize the prediction model which used only the government statistical indicators as explanatory variables in the past. To achieve this goal, this study established six types of models, including government statistical data, web search data or different combinations of them as explanatory variables.The empirical results show that the addition of web search data plays a significant role in promoting the model only use the government statistical indicators as explanatory variables. The web search data is a good complement to the government statistical data. To be noted that it should be distinguished from the government statistical data. The results of this paper can be used to help statistical work of the total retail sales of consumer goods, and also provide a reference for government departments to formulate policies to stimulate domestic demand.
In the era of big data, web search data is generated at an alarming rate every day, which enriches the type and source of data. There is still little literature involved about whether and how to use web search data to predict the total retail sales of social consumer goods. Based on the data of Hubei Province, this paper researches whether the web search data can optimize the prediction model which used only the government statistical indicators as explanatory variables in the past. To achieve this goal, this study established six types of models, including government statistical data, web search data or different combinations of them as explanatory variables.The empirical results show that the addition of web search data plays a significant role in promoting the model only use the government statistical indicators as explanatory variables. The web search data is a good complement to the government statistical data. To be noted that it should be distinguished from the government statistical data. The results of this paper can be used to help statistical work of the total retail sales of consumer goods, and also provide a reference for government departments to formulate policies to stimulate domestic demand.
引文
[1] Askitas N,Ziimmermann K F. Google Econometrics and Unemployment Forecasting[J]. Applied Economics Quarterly,2009,(2):107-120.
[2] Choi H,Varian H. Predicting Initial Claims for Unemployment Insurance Using Google Trends[R]. Technical Report,Google,2009.
[3] Suhoy T. Query Indices and a 2008 Downturn:Israeli Data[M]. Bank of Israel Discussion Papers,2009.
[4] Simeon V,Torsten S. Forecasting Private Consumption:Survey-based Indicator vs Google Trends[J]. Journal of Foresting,2011,(10):1002-1013.
[5] Vosen S,Schmidt T. Forecasting Private Consumption:Survey-Based Indicators VS. Google Trends[J]. Journal of Forecasting,2011,(6):565-578.
[6] Wu L,Brynjolfsson E. The Future of Prediction:How Google Searches Foreshadow Housing Prices and Quantities[R]. Thirtieth International Conference on Information System, 2009:1-14.
[7] Choi H,Varian H. Predicting the Present with Google Trends[R]. Technical Report,Google,2011.
[8] McLaren N,Shanbhogue R. Using Internet Search Data as Economic Indicators[J]. Bank of England Quarterly Bulletin,2011,(Q2):134-140.
[9] Schlegel G L. Utilizing Big Data and Predictive Analytics to Manage Supply Chain Risk[J]. Journal of Business Forecasting,2014,(4):11-17.
[10]宋双杰,曹晖,杨坤.投资者关注与IPO异象——来自网络搜索量的经验证据[J].经济研究,2011,(S1):145-155.
[11]赵龙凯,陆子昱,王致远.众里寻“股”千百度——股票收益率与百度搜索量关系的实证探究[J].金融研究,2013,(4):183-195.
[12]张崇,吕本富,等.网络搜索数据与CPI的相关性研究[J].管理科学学报,2012,(7):50-59.
[13]孙毅,吕本富,等.大数据视角的通胀预期测度与应用研究[J].管理世界,2014,(4):171-172.
[14]董倩,孙娜娜,李伟.基于网络搜索数据的房地产价格预测[J].统计研究,2014,(10):81-88.
[15]任乐,崔东佳.基于网络搜索数据的国内旅游客流量预测研究——以北京市国内旅游客流量为例[J].经济问题探索,2014,(4):67-73.
[16]王炼,贾建民.突发性灾害事件风险感知的动态特征:来自网络搜索的证据[J].管理评论,2014,(5):169-176.
[17]刘涛雄,徐晓飞.互联网搜索行为能帮助我们预测宏观经济吗[J].经济研究,2015,(12):68-83.
[18]李方一,肖夕林,刘思佳.基于网络搜索数据的区域经济预警研究[J].华东经济管理,2016,(8):60-66.
[19]袁铭.基于网络搜索量和混合频率模型的经济变量预测研究[J].统计与信息论坛,2016,(5):27-35.
[20]董倩.基于网络搜索数据的雾霾经济与CPI相关性研究[J].调研世界,2016,(12):58-60.
[21]董莉,等.大数据背景下的CPI实时预测研究[J].调研世界,2017,(8):51-54.
[22]耿林,毛宇飞.中国就业景气指数的构建、预测及就业形势判断[J].中国人民大学学报,2017,(6):24-35.
(1)很多经济指标一般会呈现波峰、波谷的周期性波动,一些与之相关的经济指标会在其出现波峰或波谷之前、之后抑或同时出现波峰或波谷的特征,根据出现波峰或波谷的时间将这些指标分为先行指标、滞后指标和同步指标,先行指标指的是在基准指标出现波动之前出现波动的指标。
[2] Choi H,Varian H. Predicting Initial Claims for Unemployment Insurance Using Google Trends[R]. Technical Report,Google,2009.
[3] Suhoy T. Query Indices and a 2008 Downturn:Israeli Data[M]. Bank of Israel Discussion Papers,2009.
[4] Simeon V,Torsten S. Forecasting Private Consumption:Survey-based Indicator vs Google Trends[J]. Journal of Foresting,2011,(10):1002-1013.
[5] Vosen S,Schmidt T. Forecasting Private Consumption:Survey-Based Indicators VS. Google Trends[J]. Journal of Forecasting,2011,(6):565-578.
[6] Wu L,Brynjolfsson E. The Future of Prediction:How Google Searches Foreshadow Housing Prices and Quantities[R]. Thirtieth International Conference on Information System, 2009:1-14.
[7] Choi H,Varian H. Predicting the Present with Google Trends[R]. Technical Report,Google,2011.
[8] McLaren N,Shanbhogue R. Using Internet Search Data as Economic Indicators[J]. Bank of England Quarterly Bulletin,2011,(Q2):134-140.
[9] Schlegel G L. Utilizing Big Data and Predictive Analytics to Manage Supply Chain Risk[J]. Journal of Business Forecasting,2014,(4):11-17.
[10]宋双杰,曹晖,杨坤.投资者关注与IPO异象——来自网络搜索量的经验证据[J].经济研究,2011,(S1):145-155.
[11]赵龙凯,陆子昱,王致远.众里寻“股”千百度——股票收益率与百度搜索量关系的实证探究[J].金融研究,2013,(4):183-195.
[12]张崇,吕本富,等.网络搜索数据与CPI的相关性研究[J].管理科学学报,2012,(7):50-59.
[13]孙毅,吕本富,等.大数据视角的通胀预期测度与应用研究[J].管理世界,2014,(4):171-172.
[14]董倩,孙娜娜,李伟.基于网络搜索数据的房地产价格预测[J].统计研究,2014,(10):81-88.
[15]任乐,崔东佳.基于网络搜索数据的国内旅游客流量预测研究——以北京市国内旅游客流量为例[J].经济问题探索,2014,(4):67-73.
[16]王炼,贾建民.突发性灾害事件风险感知的动态特征:来自网络搜索的证据[J].管理评论,2014,(5):169-176.
[17]刘涛雄,徐晓飞.互联网搜索行为能帮助我们预测宏观经济吗[J].经济研究,2015,(12):68-83.
[18]李方一,肖夕林,刘思佳.基于网络搜索数据的区域经济预警研究[J].华东经济管理,2016,(8):60-66.
[19]袁铭.基于网络搜索量和混合频率模型的经济变量预测研究[J].统计与信息论坛,2016,(5):27-35.
[20]董倩.基于网络搜索数据的雾霾经济与CPI相关性研究[J].调研世界,2016,(12):58-60.
[21]董莉,等.大数据背景下的CPI实时预测研究[J].调研世界,2017,(8):51-54.
[22]耿林,毛宇飞.中国就业景气指数的构建、预测及就业形势判断[J].中国人民大学学报,2017,(6):24-35.
(1)很多经济指标一般会呈现波峰、波谷的周期性波动,一些与之相关的经济指标会在其出现波峰或波谷之前、之后抑或同时出现波峰或波谷的特征,根据出现波峰或波谷的时间将这些指标分为先行指标、滞后指标和同步指标,先行指标指的是在基准指标出现波动之前出现波动的指标。