GDELT:统计特定主题的出现次数

Question

我正在尝试计算术语“比特币”在GDELT数据库的主题列中出现的频率，然后按日期对计数进行分组。这是我到目前为止所知道的：

SELECT DATE, SPLIT(RTRIM(Themes,';'),';') themes 

FROM `gdelt-bq.gdeltv2.gkg_partitioned` WHERE _PARTITIONTIME >= "2020-11-01 00:00:00" AND _PARTITIONTIME < "2020-11-07 00:00:00" 
#and (Themes like "%BITCOIN%")
#or (AllNames like "%bitcoin%" or AllNames like "%BITCOIN%")

and length(Themes) > 1
) select count(theme) cnt from nested, UNNEST(themes) as theme WHERE theme like "%BITCOIN%"

group by DATE

这是正确的方法吗？谢谢！

Answer 1

下面是针对BigQuery标准SQL的说明

#standardsql
select date(_partitiontime) date, count(theme) occurences
from `gdelt-bq.gdeltv2.gkg_partitioned`, unnest(split(themes,';')) as theme 
where _partitiontime >= "2020-11-01 00:00:00" and _partitiontime < "2020-11-07 00:00:00" 
and lower(theme) like "%bitcoin%"
group by date
-- order by date  

with output

​

​