A generalized solution to verify authorship and detect style change in multi-authored documents
Name
3625007.3627589.pdf
Size
921.72 KB
Format
Adobe PDF
Checksum (MD5)
b573fc41c6f817f8b67c0650e58cef46
Author(s)
Leekha, Rohan
Vandam, Courtland
Date Issued
November 6, 2023
Publisher
ACM
Citation
Leekha, Rohan and Vandam, Courtland. 2023. "A generalized solution to verify authorship and detect style change in multi-authored documents."
Version
Final published version
Abstract
Identifying changes in style can be used to detect multi-authored social media accounts, plagiarism, compromised accounts, and author contributions in long documents. We propose an approach to recognize changes in authorship using large language models. Our approach leverages sentence-level contextual embeddings and semantic relationships. First we expand the training set by adding adversarial examples to the minority class [5], [13], [17]. Then we fine-tune a sequence classification transformer model to detect style change. Our approach outperforms all baselines of PAN21 with macro F1-scores of 0.80, 0.74, and 0.70 for detecting style changepoint between paragraphs, closed-set author ID per paragraph, and style changepoint between sentences, respectively. Our approach also performs better than the leading competitors in PAN22. Also, we achieved a five percent improvement in macro F1-score (0.78) on the newly introduced DarkReddit+ dataset for authorship verification.
Description
ASONAM '23, November 6–9, 2023, Kusadasi, Turkiye
MIT Department
Lincoln Laboratory
Terms of Use
Creative Commons Attribution
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1145/3625007.3627589
