Sql server function to display word frequency in column

Question

I have a table that lists the input of freot text from a survey where participants were allowed to enter answers (regarding the colors that they would like to have at their wedding)

I would like to write a sql function that collects all the information from this column, and orders count the frequency of each word, ordering the result given by this counter.

Response
--------
Red and White
green
White and blue
Blue
Dark blue

I would like the table above to be arranged as follows

Response  Frequency
--------  ---------
Blue      3
White     2
And       2
Red       1
Green     1

I can delete all garbage words like "and" after the function is executed. Does anyone know of any good features that produce this behavior?

Answer 1

, . -, ...

Alter Function dbo.SeparateValues    

(    
 @data VARCHAR(MAX),    
 @delimiter VARCHAR(10)     
)     
RETURNS     
@tbldata TABLE(col VARCHAR(MAX))    
As    
--Declare @data VARCHAR(MAX) ,@delimiter VARCHAR(10)     
--Declare @tbldata TABLE(col VARCHAR(10))    
--Set @data = 'hello,how,are,you?,234234'    
--Set @delimiter = ','    
--DECLARE @tbl TABLE(col VARCHAR(10))    
Begin    
DECLARE @pos INT    
DECLARE @prevpos INT    
SET @pos = 1     
SET @prevpos = 0    

WHILE @pos > 0     
BEGIN    
SET @pos = CHARINDEX(@delimiter, @data, @prevpos+1)    
if @pos > 0     
INSERT INTO @tbldata(col) VALUES(LTRIM(RTRIM(SUBSTRING(@data, @prevpos+1, @pos-@prevpos-1))))    
else    
INSERT INTO @tbldata(col) VALUES(LTRIM(RTRIM(SUBSTRING(@data, @prevpos+1, len(@data)-@prevpos))))    
SET @prevpos = @pos     
End    

RETURN       
END    

...

Select Count(*), sep.Col FROM (
        Select * FROM (
            Select value = Upper(RTrim(LTrim(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(Replace(response, ',', ' '), '.', ' '), '!', ' '), '+', ' '), ':', ' '), '-', ' '), ';', ' '), '(', ' '), ')', ' '), '/', ' '), '&', ''), '?', ' '), '  ', ' '), '  ', ' ')))) FROM Responses
        ) easyValues
        Where value <> '' 
    ) actualValues 
    Cross Apply dbo.SeparateValues(value, ' ') sep
    Group By sep.Col
    Order By Count(*) Desc

, OTT , , .

Answer 2

, SQL Server split.

, .

http://www.sqlteam.com/forums/topic.asp?TOPIC_ID=50648

, proc ...

CREATE TABLE #Temp (Response nvarchar(50), Frequency int) 

DECLARE @response nvarchar(100)
DECLARE db_cursor CURSOR FOR 
SELECT response FROM YourTable

OPEN db_cursor  
FETCH NEXT FROM db_cursor INTO @response 

WHILE @@FETCH_STATUS = 0  
BEGIN  
       /* Pseudo Code */ 
       --Split @Response 
       --Iterate through each word in returned list
       --IF(EXISTS in #TEMP)
       --    UPDATE THAT ROW & INCREMENT THE FREQUENCY
       --ELSE
       --    NEW WORD, INSERT TO #Temp WITH A FREQUENCY OF 1

       FETCH NEXT FROM db_cursor INTO @response 
END   

SELECT * FROM #Temp

Theres, , , -, , ,

Answer 3

DECLARE @phrases TABLE (id int, phrase varchar(max))
INSERT @phrases values
(1,'Red and White'  ),
(2,'green'          ),
(3,'White and blue' ),
(4,'Blue'           ),
(5,'Dark blue'      );

SELECT word, COUNT(*) c
FROM @phrases
CROSS APPLY (SELECT CAST('<a>'+REPLACE(phrase,' ','</a><a>')+'</a>' AS xml) xml1 ) t1
CROSS APPLY (SELECT n.value('.','varchar(max)') AS word FROM xml1.nodes('a') x(n) ) t2
GROUP BY word

word         freq
----------- -----------
and         2
blue        3
Dark        1
green       1
Red         1
White       2