Summary of Towards Robustness Of Text-to-visualization Translation Against Lexical and Phrasal Variability, by Jinwei Lu et al.
Towards Robustness of Text-to-Visualization Translation against Lexical and Phrasal Variability
by Jinwei Lu, Yuanfeng Song, Haodi Zhang, Chen Zhang, Raymond Chi-Wing Wong
First submitted to arxiv on: 10 Apr 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
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 Text-to-Vis is an emerging natural language processing (NLP) task that generates data visualizations from natural language questions. Current models rely heavily on lexical matching, which may compromise robustness against input variations. This study examines the robustness of existing text-to-vis models using a new dataset, nvBench-Rob, which contains diverse lexical and phrasal variations based on the original benchmark, nvBench. Results show that existing models’ performance drops dramatically on this new dataset, indicating inadequate robustness overall. To address input perturbations, the authors propose GRED, a novel framework combining Retrieval-Augmented Generation (RAG) with three components: NLQ-Retrieval Generator, Visualization Query-Retrieval Retuner, and Annotation-based Debugger. Experimental evaluations demonstrate that GRED outperforms RGVisNet in terms of model robustness, achieving a 32% increase in accuracy on the proposed dataset. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine being able to ask a computer a question and having it create a helpful graph or chart to answer your question. This is called text-to-vis, and it’s an important area of research in natural language processing (NLP). Right now, most text-to-vis models are not very good at handling different ways people might ask the same question. This study looked at how well these models do when faced with unexpected questions or variations. The researchers found that current models don’t do very well and proposed a new approach called GRED to improve their performance. |
Keywords
» Artificial intelligence » Natural language processing » Nlp » Rag » Retrieval augmented generation
