t**********r
发帖数: 182 | 1 Here I have a large database. One key variable is BIO (CEO/CFO's bio
description). I want to find the words that are mostly frequently appearned;
say, the first 100 words that are mostly used. How can I do this using SAS?
many thanks. | A*******s
发帖数: 3942 | 2 SAS EM有个text miner。虽然我没用过，不过顾名思义+看图识字，应该是它了吧。
appearned;
SAS?
【在 t**********r 的大作中提到】
: Here I have a large database. One key variable is BIO (CEO/CFO's bio
: description). I want to find the words that are mostly frequently appearned;
: say, the first 100 words that are mostly used. How can I do this using SAS?
: many thanks.
| d*******1
发帖数: 854 | 3 data bioword;
set database;
i=1;
do while (scan(bio,i,' ') ne '') ;
word=scan(bio,i,' ');
i=i+1;
output;
end;
keep word;
run;
proc freq data=bioword;
table word/out=freq;
run;
proc sort data=freq; by descending count; run;
appearned;
SAS?
【在 t**********r 的大作中提到】
: Here I have a large database. One key variable is BIO (CEO/CFO's bio
: description). I want to find the words that are mostly frequently appearned;
: say, the first 100 words that are mostly used. How can I do this using SAS?
: many thanks.
| A*******s
发帖数: 3942 | 4 鸡蛋里挑挑骨头--还得考虑uppercase/lowercase, single/plural,tenses for verbs...
【在 d*******1 的大作中提到】
: data bioword;
: set database;
: i=1;
: do while (scan(bio,i,' ') ne '') ;
: word=scan(bio,i,' ');
: i=i+1;
: output;
: end;
: keep word;
: run;
|
|
