FPT: Feature Prompt Tuning for Few-shot Readability Assessment
This work addresses the challenge of few-shot readability assessment for natural language processing applications, offering a novel tuning framework that integrates linguistic knowledge, though it is incremental in advancing prompt-based methods for a specific domain.
The paper tackles the problem of few-shot readability assessment by proposing Feature Prompt Tuning (FPT), a method that incorporates linguistic features into soft prompts and uses a new loss function, resulting in significant performance improvements over prior prompt-based and linguistic-feature methods, and outperforming GPT-3.5-turbo-16k in most cases.
Prompt-based methods have achieved promising results in most few-shot text classification tasks. However, for readability assessment tasks, traditional prompt methods lackcrucial linguistic knowledge, which has already been proven to be essential. Moreover, previous studies on utilizing linguistic features have shown non-robust performance in few-shot settings and may even impair model performance.To address these issues, we propose a novel prompt-based tuning framework that incorporates rich linguistic knowledge, called Feature Prompt Tuning (FPT). Specifically, we extract linguistic features from the text and embed them into trainable soft prompts. Further, we devise a new loss function to calibrate the similarity ranking order between categories. Experimental results demonstrate that our proposed method FTP not only exhibits a significant performance improvement over the prior best prompt-based tuning approaches, but also surpasses the previous leading methods that incorporate linguistic features. Also, our proposed model significantly outperforms the large language model gpt-3.5-turbo-16k in most cases. Our proposed method establishes a new architecture for prompt tuning that sheds light on how linguistic features can be easily adapted to linguistic-related tasks.