SQL Server Aggregate using like - sql

Question

I have a table with 2 columns, one is a 1/0 flag for whether they've opened an email, the second is the email address (ie joe123#domain.com).
Opened Email
0 joe123#domain.com
1 sue234#email.net
... ...
I'm trying to find if certain patterns in user names affect open rates by using regex patterns in LIKE but am unsure of the syntax for rows that match AND do NOT match the pattern.
For instance, I can do:
SELECT Email, sum(Opened)
FROM table1
WHERE Email LIKE '%joe%'
But this only gives me rows that match. I'd like rows that do and DON'T match in the same output.
I'd like to get something like this:
Pattern Opened
'joe' 55
not_joe 15987
'sue' 78
not_sue 15964
... ...
What's the best way to do this?

Answer 1

If you've already got a list of patterns, you can achieve the LIKE/NOT LIKE by using a CROSS JOIN (warning: performance hit).
See below for an example. Note: you can potentially improve the performance of the LIKE in the SELECT statement - there are other options available.
DECLARE #MatchList TABLE (ID INT, Pattern VARCHAR(3))
INSERT INTO #MatchList (ID, Pattern) VALUES (1, 'Joe')
INSERT INTO #MatchList (ID, Pattern) VALUES (2, 'Sue')
DECLARE #Table1 TABLE (Email VARCHAR(100))
INSERT INTO #Table1 (Email) VALUES ('joe123#domain.com')
INSERT INTO #Table1 (Email) VALUES ('sue234#email.net')
INSERT INTO #Table1 (Email) VALUES ('sue682#email.net')
INSERT INTO #Table1 (Email) VALUES ('a#domain.com')
INSERT INTO #Table1 (Email) VALUES ('b#domain.com')
INSERT INTO #Table1 (Email) VALUES ('c#domain.com')
INSERT INTO #Table1 (Email) VALUES ('d#domain.com')
INSERT INTO #Table1 (Email) VALUES ('e#domain.com')
SELECT
CASE WHEN Email LIKE '%'+Pattern+'%' THEN Pattern ELSE 'not_'+Pattern END AS Classification,
COUNT(*) Opened
FROM
#MatchList m
CROSS JOIN
#Table1
GROUP BY m.ID, CASE WHEN Email LIKE '%'+Pattern+'%' THEN Pattern ELSE 'not_'+Pattern END
ORDER BY m.ID

Answer 2

If i understood correctly you are trying to identify similar usernames
you can use a function like SOUNDEX to group your count by
SELECT
Email,
SOUNDEX(Email)
FROM TableA
SELECT COUNT(*) OpenedCount,
SOUNDEX(Email) EmailSOUNDEX
FROM TableA
GROUP BY SOUNDEX(Email)
ORDER BY OpenedCount desc
If all you want is something like all emails that have sue in it?
SELECT COUNT(*) OpenedCount,
SOUNDEX(Email) EmailSOUNDEX
FROM TableA
WHERE
Email like '%Sue%'
GROUP BY SOUNDEX(Email)
ORDER BY OpenedCount desc
For this kind of search maybe you can build a dictionary with relevant search terms and join it with
SELECT COUNT(*) OpenedCount,
SearchTermTable.Term
FROM TableA
JOIN SearchTermTable ON
TableA.Email like '%'+SearchTermTable.Term+'%'
WHERE
GROUP BY SearchTermTable.Term
ORDER BY OpenedCount desc