• S. Ashok Kumar
• D. Arvind
• S. Dharanikumar
• S. Premkumar

Many software and applications are being developed on a daily basis; millions of lines of code are being written daily, thousands and thousands of bugs are fixed daily. Bugs are the common enemies for developers, the larger the code base of software the more troublesome they are. The possibility is very high that a bug. Bug fixing normally takes a lot of human resources and time to fix it. A developer is said to spend a substantial amount of time in fixing a bug. Though there are sites like Stack Overflow, a developers community which is like an encyclopaedia for developers, doesn’t help the developers in finding the right answers all the time, and mostly the amount of time a developer spends in finding a fix for the bugs are a lot. So a tool which can suggest a fix for the bugs in the program for the programmers can be of a great value. The easiest way is to manually record the bugs and fix used for it, by classifying them can make much sense but it is a tedious process to save the bug and the fix used and finding the right fix may take a solid amount of time So we are suggesting a model, BUGCHER which can learn the program transformations of code which are done by some programmer to fix an issue by producing a patch for a bug. The model takes the line which has the bug as input and produces a fix which may be able to fix the bug with some tweaks. And another use case of our idea is to make a automotive grading system for programming tasks. It is not a cake walk thing to grade or correct the mistakes in the programming tasks by single individual, the more the number of students the more the time it takes for the tutors to grade the assignments. Moreover the tutor can’t correct the mistakes and be there for all the students to help them find the right answer. Thus we are suggesting a way by which we can use our model to help the students. The approach is, at first the tutor feeds our model with possible mistakes the students may make in their assignments and the corresponding fix for it. These sets of mistakes (i.e. bugs) and their fixes can be retrieved from previous assignments and also from the tutor’s years of experience.

Automatic Patch Generation

"Automatic Patch Generation System Using Deep Learning Techniques", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 9, page no.949-953, September 2020, Available :http://www.jetir.org/papers/JETIR2009426.pdf

"Automatic Patch Generation System Using Deep Learning Techniques", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 9, page no. pp949-953, September 2020, Available at : http://www.jetir.org/papers/JETIR2009426.pdf