【摘要】 [目的/意义]我国有关大数据分析的研究中流行一些说法:数据能讲述故事;要全体不要抽样;要相关不要因果。这些说法可能具有误导性。[方法/过程]通过分析纽约市"市长数据分析办公室"利用大数据分析帮助改善城市部门绩效方面的成功案例,检验上述说法的正确性。[结果/结论]大数据分析不等于让数据自己说话,数据自己是不会说话的。成功的大数据分析首先需要分析人员确定方向、目标,然后根据它们确定分析的对象,有针对性的收集和分析数据。数据够了就好,"要全体不要抽样"不是必要的。大数据分析首先需要了解数据的背景,离开了其所处的背景的数据是没有价值的。大数据分析除了关注数据,还需要关注数据之外的很多因素,包括政治因素。更多还原

【Abstract】 [Purpose/Significance]There are some popular arguments that may be misleading in China’s research on big data analysis such as:the data themselves can tell the story;using all the data available,not just sample sets;having relevance,not causation.[Methods/Process]This paper analyzes the successful practice of the NewYork City’s"Mayor’s Office of Data Analysis"which uses big data analysis to help improve the performance of the government agencies of the city and tests the above arguments.[Results/Conclusion]The conclusions are that the big data analysis does not mean that the data can speak for themselves,data cannot speak by themselves;big data analysis first requires analysts to determine the direction and goals of the analysis,then to determine the objects of analysis according to the direction and goals and to carry out the targeted data collection and analysis;as data scale is concerned,enough is enough,it is not necessary"to have all data";it is necessary to understand the background of the data before data analysis,data without its background is of no value;more factors other than data should be considered in big data analysis,including political factors.更多还原