For example, a teacher could use a good plagiarism checker as Unplag or Turnitin and check papers for possible similarities. These automated software helps to check against its huge database (Turnitin) and against the Internet (Unplag) and receive the results in minutes! However, I believe that the final grade is up to the teacher. As a human he or she can evaluate student's paper in a more precise way, using the results of automated check performed
by the program.
Automated grading software is useful for checking to see if plagiarism is being used. There is no way for a teacher to check or a paper against thousands or millions of possible copies, but a computer program can. A human would still be needed for the finer points of grading such as the paper being topical to the assignment, but automated grading can help speedup the process, and fill in holes in a human only grading system.
Machines simply don't have the flexibility needed to grade something as subjective as a paper. It is a good idea to run papers through a plagiarism detection program, but grading? No way. Things like style and how well an argument is constructed could not be properly graded by a computer program.
Papers and essays should not be graded by automated software. Technology makes mistakes and can understand what is beyond the words on a piece of paper. It would only be able to grade spelling and grammar, and essays have much more to offer than just that. Students put effort into their work, they deserve to have it read and looked at by a teacher or professor.
Automated software is not even close to being able to break down the English language enough for this to be a reasonable grading system. They cannot understand context, style or even things as simple as word choices yet. This will probably be reality some day but we are not close to it.