Plagiarism Detection in Programming Assignments Using Deep Features

Abstract

This paper proposes a method for detecting plagiarism in source-codes using deep features. The embeddings for programs are obtained using a character-level Recurrent Neural Network (char-rnn), which is pre-trained on Linux Kernel source-code. Many popular plagiarism detection tools are based on n-gram techniques at syntactic level. However, these approaches to plagiarism detection fail to capture long term dependencies (non-contiguous interaction) present in the source-code. Contrarily, the proposed deep features capture non-contiguous interaction within n-grams. These are generic in nature and there is no need to fine-tune the char-rnn model again to program submissions from each individual problem-set. Our experiments show the effectiveness of deep features in the task of classifying assignment program submissions as copy, partial-copy and non-copy. Comparing our proposed features with handcrafted features (source-code metrics and textual features), we report f1-score improvement of 9.5% for binary classification and 5% for three-way classification tasks respectively.