Analyzing a Large Corpus of Crowdsourced Plagiarism Master’s Thesis