Best approach to compare rows of table(Java/Oracle)

I have a HISTORY table with million rows and having 300 columns. For one unique ID there could be max of 5000 rows which I want to compare like

Row1 with Row2, Row2 with Row3, Row3 with Row4, . . . Row4999 with Row5000

and find out the data differences in the columns and the output should be something like:

What are the implications of processing the rows on Java side? or is it better to process on the DB side and push the output to Java?

from Newest questions tagged java - Stack Overflow http://ift.tt/1MuMyyr
via IFTTT