Is this the end of the gold standard? A straightforward reference-less grammatical error correction metric

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 257–268 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 257 Progressive Self-Training with Discriminator for Aspect Term Extraction Qianlong Wang1, Zhiyuan Wen1, Qin Zhao1, Min Yang2, Ruifeng Xu1,3∗ 1Harbin Institute of Technology (Shenzhen), China 2Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences 3Peng Cheng Laboratory, Shenzhen, China qlwang15@outlook.com, wenzhiyuan2012@gmail.com zhaoqin@hit.edu.cn min.yang@siat.ac.cn, xuruifeng@hit.edu.cn Abstract Aspect term extraction aims to extract aspect terms from a review sentence that users have expressed opinions on. One of the remain- ing challenges for aspect term extraction re- sides in the lack of sufficient annotated data. While self-training is potentially an effective method to address this issue, the pseudo-labels it yields on unlabeled data could induce noise. In this paper, we use two means to alleviate the noise in the pseudo-labels. One is that in- spired by the curriculum learning, we refine the conventional self-training to progressive self-training. Specifically, the base model in- fers pseudo-labels on a progressive subset at each iteration, where samples in the subset be- come harder and more numerous as the iter- ation proceeds. The other is that we use a discriminator to filter the noisy pseudo-labels. Experimental results on four SemEval datasets show that our model significantly outperforms the previous baselines and achieves state-of- the-art performance. 1 Introduction Aspect term extraction (ATE) is a crucial task in aspect-level sentiment analysis, aiming to extract all aspect terms present in the sentence (Pontiki et al., 2014). For example, given a restaurant re- view "I looove their eggplant pizza, as well as their pastas!", ATE system aims to extract "eggplant pizza" and "pastas". Many researchers typically formulated ATE as a sequence labeling problem or a token-level classi- fication problem. The current state-of-the-art neu- ral models can be classified into two categories. One designs the sophisticated model with a variety of techniques, such as history attention (Li et al., 2018), sequence to sequence (Ma et al., 2019), and constituency lattice (Yang et al., 2020). Although these models achieve satisfactory performance, its sufficient condition is the availability of sufficient ∗ Corresponding author. training data. However, labeling a large amount of aspect data may not be practical due to its cost. The other aims at addressing the data insuffi- ciency issue from a different perspective. For ex- ample, Li et al. (2020) generated the reviews while preserving the original aspects via formulating the data augmentation as a conditional generation task. However, varying only a small number of consec- utive non-aspect words in the reviews would limit the semantic diversity of the sample. Chen and Qian (2020) tackled long-tail distributions problem of aspect terms and context words in the training sets with soft prototypes. Nevertheless, the way that the soft template implicitly uses external data discounts the usefulness of external data. Different from previous approaches, in this pa- per, we use self-training (Scudder, 1965) to allevi- ate the labeled data insufficiency. In self-training, a base model trained on the labeled data is used to …

Content truncated to 500 words