The National Institution for Academic Degrees and Quality Enhancement of Higher Education (NIAD-QE) offers a bachelor’s degree conferment program for graduates of junior colleges, colleges of technology, and professional training colleges based on accumulated credits. The authors have mainly conducted research on a support system for judging the classification of courses, which is important when judging the awarding of bachelor’s degrees using this system. On the other hand, it is placing an increasing burden not only on the classification of courses, but also on the faculty review. In this study, we focus on the examination work related to faculty review, and discuss the possibility of using deep learning to support work related to degree awarding on NIAD-QE. In particular, we consider verifying similarities between documents using word2vec and reducing the workload involved in conducting a short thesis examination using large-scale language models.