QGIS 3x - categorized style customization based on a part of the string

Assuming Consistent Formatting in TITLE Field

The expression substr("TITLE", 4, 1) returns the fourth character of the string. If the TITLE field is consistently formatted, this will correspond to the particular category, either 1 or 2. Just enter that expression and click classify.

Alternatively, you could modify substr's starting position and length, or if the SE-# is always at the beginning, simply use left("TITLE", 4) to get the SE-# text, not just the number.

Inconsistent Formatting in TITLE Field

If the SE-# text does not consistently appear at the beginning of the string, you can use the regex_substr function. This is easiest for grabbing the entire SE-# text, but can be modified to grab just the number, or you can nest it within a right function instead.

regexp_substr("TITLE", 'SE-[1,2]')

Have you tried Conditional expressions like IF or CASE? Something like if( "TITLE" = 'SE-1' , 1, 0 )

In addition to Cyrmel: The formula is "TITLE" LIKE '%SE-2%' and than you have to classify (to 1/0 or true/false). If you want, you can also change manually the 0 and 1 in the legend.

You did not specify the value which triggers whether your polygon should go into the first group or not - "TITLE" LIKE %SE-2% returns 1 for true and 0 for false. instead of "all other Values", set one colour to 1 and one to 0.