Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Knowledgeable Prompt-tuning: Incorporating Knowledge into Prompt Verbalizer for Text Classification

About

Tuning pre-trained language models (PLMs) with task-specific prompts has been a promising approach for text classification. Particularly, previous studies suggest that prompt-tuning has remarkable superiority in the low-data scenario over the generic fine-tuning methods with extra classifiers. The core idea of prompt-tuning is to insert text pieces, i.e., template, to the input and transform a classification problem into a masked language modeling problem, where a crucial step is to construct a projection, i.e., verbalizer, between a label space and a label word space. A verbalizer is usually handcrafted or searched by gradient descent, which may lack coverage and bring considerable bias and high variances to the results. In this work, we focus on incorporating external knowledge into the verbalizer, forming a knowledgeable prompt-tuning (KPT), to improve and stabilize prompt-tuning. Specifically, we expand the label word space of the verbalizer using external knowledge bases (KBs) and refine the expanded label word space with the PLM itself before predicting with the expanded label word space. Extensive experiments on zero and few-shot text classification tasks demonstrate the effectiveness of knowledgeable prompt-tuning.

Shengding Hu, Ning Ding, Huadong Wang, Zhiyuan Liu, Jingang Wang, Juanzi Li, Wei Wu, Maosong Sun• 2021

Related benchmarks

TaskDatasetResultRank
Sentiment ClassificationMR (test)
Accuracy86.8
142
Sentiment AnalysisSST-2 (test)
Accuracy86.8
136
Text ClassificationAGNews
Accuracy86.7
119
Text ClassificationDBpedia (DBP)
Accuracy87.4
110
Text ClassificationIMDB
Accuracy94
107
Topic ClassificationAG News (test)
Accuracy84.8
98
Topic ClassificationDBPedia (test)
Accuracy82.2
64
Sentiment ClassificationCR (test)
Mean Accuracy86.7
58
Natural Language InferenceRTE (test)
Accuracy65.8
52
Paraphrase DetectionQQP (test)
Accuracy71.6
51
Showing 10 of 22 rows

Other info

Code

Follow for update