Summary of Authorship Verification Based on the Likelihood Ratio Of Grammar Models, by Andrea Nini et al.
Authorship Verification based on the Likelihood Ratio of Grammar Models
by Andrea Nini, Oren Halvani, Lukas Graner, Valerio Gherardi, Shunichi Ishihara
First submitted to arxiv on: 13 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed method for authorship verification calculates a quantity called λ_G (LambdaG), which measures the likelihood of a document given a model of the Grammar for the candidate author, compared to a reference population. This approach uses n-gram language models trained solely on grammatical features and outperforms established methods like fine-tuned Siamese Transformer networks in terms of accuracy and AUC. The empirical evaluation demonstrates LambdaG’s robustness to genre variations and ease of interpretation, making it a promising alternative to current state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Authorship Verification is the process of identifying whether documents were written by a specific person. This problem often arises in forensic cases where documents are used as evidence. Existing methods use computational solutions that aren’t supported by scientific explanations and can be difficult for analysts to understand. To address this, researchers propose a new method that calculates a quantity called λ_G (LambdaG). This approach uses language models trained on grammatical features and outperforms other established methods in terms of accuracy and ease of interpretation. |
Keywords
* Artificial intelligence * Auc * Likelihood * N gram * Transformer
