Summary of Transparent Neighborhood Approximation For Text Classifier Explanation, by Yi Cai et al.
Transparent Neighborhood Approximation for Text Classifier Explanation
by Yi Cai, Arthur Zimek, Eirini Ntoutsi, Gerhard Wunder
First submitted to arxiv on: 25 Nov 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 paper introduces a new approach for generating model-agnostic explanations in text classification tasks. The authors highlight the limitations of current methods, which rely on neural networks and lack transparency. To overcome this limitation, they propose a probability-based editing method that generates neighboring texts by manipulating contextual information within the text. This approach, called XPROB, achieves competitive performance compared to existing generator-based methods on two real-world datasets while providing a fully transparent and controllable construction process. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper explains how neighborhood construction helps with model-agnostic explanations in text classification. It shows that current methods are not very transparent and introduces a new way to generate neighboring texts using probability editing. This makes the explanation process more stable and controlled. |
Keywords
» Artificial intelligence » Probability » Text classification
