A Study on AI-based SW Similarity Detection Model Design

Abstract

Currently, we live in an era called the Fourth Industrial Revolution, and as a key technology representing this era, we use artificial intelligence instead of computers for most of the intelligence activities that can be performed using the human brain. Various systems and SW are gradually evolving due to artificial intelligence, and SW usability is increasing very rapidly. However, as SW usability increases, problems such as reproduction and misuse occur, and research institutes and companies are demanding SW similarity. Accordingly, this study aims to design a similarity detection model using artificial intelligence for SW replication emotion. The similarity detection model design of this study is based on the Few-Shot Learning in a similarity detection model proceeds with meta-learning using DataSet with sufficient data, the data with less data contained in each class, and the recognition and learning of deep learning that rationally expresses the human brain.