Ziyu Liu

Master's

Computer Science

HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY

Software Engineer Intern

Utilized advanced web scraping techniques and network requests, developed a script to crawl and retrieve Java code datasets from Google Code Jam Applied TF-IDF algorithm and N-Gram model to assess word frequency and extract keywords; utilized the Tree-sitter tool to transform the code into abstract syntax tree, obtain the graph structure information of the code Built a model using LSTM with the Adam optimizer to train the extracted feature vectors, the proposed model achieved an accuracy rate of 95% in analyzing the authorship similarity for the labeled Java dataset; utilized to detect plagiarism and code reuse in students’ projects, ensuring integrity and fostering genuine innovation and originality