Leveraging LLM and RAG for Automated Answer Script Evaluation | IEEE Conference Publication | IEEE Xplore