Public data accompanying "The relationship between the human needs of places and economic conditions in America's 21 largest cities"

This repository contains two datasets: category_need_vector.csv and zipcode_human_need.csv <strong>category_need_vector.csv: </strong>For each of 1025 Yelp POI types (i.e. categories), we provide the percent of respondents who listed this POI as: need_1, need_2, need_3, need_4 or need_5. This data was gathered through Mechanical Turk. We also provide the top level categories for each POI type. <strong>zipcode_human_need.csv: </strong>For zip codes (n = 1186) in 21 selected US cities, we provide the z score of each need, calculated for the aggregate of the POIs within the zip code. The z score is centered around 0 as a mean, and -1 or 1, as one standard deviation in either direction, -2 or 2 as two standard deviations in either direction, etc. We also provide the entropy of each zip code, where higher numbers indicate more deviation between the needs. <br>

本仓库包含两个数据集：category_need_vector.csv 与 zipcode_human_need.csv。
<strong>category_need_vector.csv：</strong> 针对1025种Yelp平台兴趣点（Point of Interest，POI）类型（即品类），我们提供了将该兴趣点列为need_1、need_2、need_3、need_4或need_5的受访者占比。该数据通过机械 Turk（Mechanical Turk）平台采集获得，同时我们还提供了每种兴趣点类型对应的一级品类分类。
<strong>zipcode_human_need.csv：</strong> 针对美国21个选定城市中的1186个邮政编码区域，我们提供了各需求的z分数，该分数基于对应邮政编码区域内所有兴趣点的聚合数据计算得到。该z分数以0为均值，±1分别对应正负一个标准差，±2对应正负两个标准差，以此类推。此外我们还提供了每个邮政编码区域的熵值，熵值越高则代表该区域内各类需求的差异程度越大。