Deep Context- and Relation-Aware Learning for Aspect-based Sentiment Analysis

Oh, Shinhyeok; Lee, Dongyub; Whang, Taesun; Park, IlNam; Seo, Gaeun; Kim, EungGyun; Kim, Harksoo

doi:10.18653/v1/2021.acl-short.63

Cited by 15 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Syntactic dependency between an aspect and its corresponding opinion expression has also been explored [13,18,24]. By utilizing additional syntactic knowledge obtained from external syntax parsers, the relative position in a syntactic tree is used to measure the distance between aspect-related terms and opinion-bearing text span in the sentence.…”

Section: Related Workmentioning

confidence: 99%

Aspect-based Sentiment Analysis through EDU-level Attentions

Lin¹,

Sun²,

Wang³

2022

Preprint

View full text Add to dashboard Cite

A sentence may express sentiments on multiple aspects. When these aspects are associated with different sentiment polarities, a model's accuracy is often adversely affected. We observe that multiple aspects in such hard sentences are mostly expressed through multiple clauses, or formally known as elementary discourse units (EDUs), and one EDU tends to express a single aspect with unitary sentiment towards that aspect. In this paper, we propose to consider EDU boundaries in sentence modeling, with attentions at both word and EDU levels. Specifically, we highlight sentiment-bearing words in EDU through word-level sparse attention. Then at EDU level, we force the model to attend to the right EDU for the right aspect, by using EDU-level sparse attention and orthogonal regularization. Experiments on three benchmark datasets show that our simple EDU-Attention model outperforms state-of-the-art baselines. Because EDU can be automatically segmented with high accuracy, our model can be applied to sentences directly without the need of manual EDU boundary annotation.

show abstract

Section: Related Workmentioning

confidence: 99%

Aspect-based Sentiment Analysis through EDU-level Attentions

Lin¹,

Sun²,

Wang³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Venugopalan et al [41] proposed a guided LDA model with BERT for each aspect category. Oh et al [42] proposed a deep contextualized relation-aware network (DCRAN) for aspect extraction. They designed two modules to capture the association relationship between subtasks of ABAE with the contextualized information.…”

Section: Related Workmentioning

confidence: 99%

Aspect term extraction via information-augmented neural network

Liu

Shen

2022

Complex Intell. Syst.

View full text Add to dashboard Cite

Aspect term extraction (ATE) aims at identifying the aspect terms that are expressed in a sentence. Recently, Seq2Seq learning has been employed in ATE and significantly improved performance. However, it suffers from some weaknesses, such as lacking the ability to encode the more informative information and integrate information of surrounding words in the encoder. The static word embeddings employed in ATE fall short of modeling the dynamic meaning of words. To alleviate the problems mentioned above, this paper proposes the information-augmented neural network (IANN) which is a novel Seq2Seq learning framework. In IANN, a specialized neural network is developed as the key module of the encoder, named multiple convolution with recurrence network (MCRN), to encode the more informative information and integrate information of surrounding words in the encoder. The contextualized embedding layer is designed to capture the dynamic word sense. Besides, the novel AO ({Aspect, Outside}) tags are proposed as the less challenging tagging scheme. A lot of experiments have been performed on three widely used datasets. These experiments demonstrate that the proposed IANN acquires state-of-the-art results and validate that the proposed IANN is a powerful method for the ATE task.

show abstract

“…Extracting emotion causes help to understand what emotions are expressed in the text and why these emotions are perceived. The emotion cause analysis can be applied in many fields, such as improving service quality [9]- [11], controlling public opinion trends [12]- [15] and assisting the empathetic response generation [16], [17]. Particularly, emotion cause extraction (ECE) aims to determine which clauses contain the causes for the given emotion [18]- [20].…”

Section: Introductionmentioning

confidence: 99%

Emotion Prediction Oriented Method With Multiple Supervisions for Emotion-Cause Pair Extraction

Zhao

2023

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Emotion-cause pair extraction (ECPE) task aims to extract all the pairs of emotions and their causes from an unannotated emotion text. The previous works usually extract the emotion-cause pairs from two perspectives of emotion and cause. However, emotion extraction is more crucial to the ECPE task than cause extraction. Motivated by this analysis, we propose an end-to-end emotion-cause extraction approach oriented toward emotion prediction (EPO-ECPE), aiming to fully exploit the potential of emotion prediction to enhance emotion-cause pair extraction. Considering the strong dependence between emotion prediction and emotion-cause pair extraction, we propose a synchronization mechanism to share their improvement in the training process. That is, the improvement of emotion prediction can facilitate the emotion-cause pair extraction, and then the results of emotion-cause pair extraction can also be used to improve the accuracy of emotion prediction simultaneously. For the emotion-cause pair extraction, we divide it into genuine pair supervision and fake pair supervision, where the genuine pair supervision learns from the pairs with more possibility to be emotion-cause pairs. In contrast, fake pair supervision learns from other pairs. In this way, the emotion-cause pairs can be extracted directly from the genuine pair, thereby reducing the difficulty of extraction. Experimental results show that our approach outperforms the 13 compared systems and achieves new state-of-the-art performance.

show abstract

Deep Context- and Relation-Aware Learning for Aspect-based Sentiment Analysis

Cited by 15 publications

References 28 publications

Aspect-based Sentiment Analysis through EDU-level Attentions

Aspect-based Sentiment Analysis through EDU-level Attentions

Aspect term extraction via information-augmented neural network

Emotion Prediction Oriented Method With Multiple Supervisions for Emotion-Cause Pair Extraction

Contact Info

Product

Resources

About