With China's national strength enhanced, the national economy, population, resources, environment and so on various aspects of the census, more and more, but because takers are all personnel at the grass-roots level, and the general survey time is short, even with training and layers of the training, lead to grassroots takers received census and eventually census data summary analysis personnel requirements, and different regional census personnel to fill in the census form is different, this needs to analyze the census data, screening and washing, how to use the computer quickly help enumerators to quickly screen out outliers? We all talk about how to deal with,

CodePudding user response:

Although I am not make census, but this is all too common: data collection, this cannot be completely avoided, usually situation is as follows:
1) although the format is not standard but are relatively accurate, big data processing of unstructured data easily calm,
2) the Data is not accurate, the accuracy can not be tolerated, or need to primary screening, for example you can design a Scan to Data software to help you troubleshoot, even can identify human face, the little Case,