The following approach builds a result of 1000 integers, then uses those integers (n) to locate segments within the comma seperated string, and for each segment it creates a new row so that the derived table looks like this:
userid | book
:----- | :---------
ym0001 | dictionary
ym0002 | textbooks
ym0001 | textbooks
ym0002 | dictionary
ym0001 | notebooks
Once that exists it is a simple matter of grouping by book to arrive at the counts.
select
book, count(*) Counts
from (
select
t.userid
, SUBSTRING_INDEX(SUBSTRING_INDEX(t.books, ',', numbers.n), ',', -1) book
from (
select @rownum:=@rownum+1 AS n
from
(
select 0 union all select 1 union all select 2 union all select 3
union all select 4 union all select 5 union all select 6
union all select 7 union all select 8 union all select 9
) a
cross join (
select 0 union all select 1 union all select 2 union all select 3
union all select 4 union all select 5 union all select 6
union all select 7 union all select 8 union all select 9
) b
cross join (
select 0 union all select 1 union all select 2 union all select 3
union all select 4 union all select 5 union all select 6
union all select 7 union all select 8 union all select 9
) c
cross join (select @rownum:=0) r
) numbers
inner join mytable t
on CHAR_LENGTH(t.books)
-CHAR_LENGTH(REPLACE(t.books, ',', '')) >= numbers.n-1
) d
group by
book
order by
book
book | Counts
:--------- | -----:
dictionary | 2
notebooks | 1
textbooks | 2
- If you already have a table of numbers, use that instead.
- the cross joins of a b and c dynamically produce 1000 rows, if you need more add further cross joins similar to c. i.e. the number of numbers should exceed the maximum length of your comma seperated data
db<>fiddle here