Code Assessment Of Student Java Program Using Static Approach

The analysis of programming assignment manually is difficult,error prone and time consuming task when it comes to huge number of students enrolled in programming courses.The solution to such problem is to have automatic grading system which assist the student assignments electronically and improve student programming learning skills by more practices.This thesis presents and automatic grading approach for java programming assignment of students named “online coding ghost”.It is based on web application that facilitate both instructor and the student to submit assignments and check evaluation results generated by the system.The assignment grading evaluation is performed by comparing the similarity between student program and each solution model programs after they both are being normalized and transformed into intermediate representation called abstract syntax tree.Teaching and learning programming is tedious task when huge number of students enrolled in programming courses.To overcome this problem and make learning process more convenient for students as well as instructors we developed an automatic grading system for programming assessment.It evaluate student program using static analysis approach.The semantic similarity between student and model program is calculated on the basis of statements,structure and size of program.System provide reasonable grading along with feedback channel which makes the learning process more robust.The application facilitates the students to online submit their programming assignments along with feedback channel.The instructor can run and test the student source code and compare it with number of model programs in order to get accuracy of results generated by the system.The assessment system produce more objective results as compare to manually because it is done by machine.This approach is applicable for object oriented languages such as c++,c# and java.However we examined this for java programming assignments.