Papers with Aspect-level sentiment analysis
A Position-aware Bidirectional Attention Network for Aspect-level Sentiment Analysis (C18-1)
Copied to clipboard
| Challenge: | Existing approaches to model aspect-level sentiment analysis neglect the position information of aspect terms. |
| Approach: | They propose a position-aware bidirectional attention network (PBAN) that mutually models the relation between aspect terms and sentences by employing bidirectional mechanism. |
| Outcome: | The proposed model can distinguish sentiment polarity of aspect terms in sentences . it can also model relation between aspect terms and sentences based on position information . |
Enhancing Aspect-level Sentiment Analysis with Word Dependencies (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing approaches to enhance aspect-level sentiment analysis have omitted syntactic information . experimental results show that our approach outperforms baseline models on all datasets . |
| Approach: | They propose to leverage word dependencies to enhance aspect-level sentiment analysis . they propose to use key-value memory networks to leverage different dependency results . |
| Outcome: | The proposed approach outperforms baseline models on all datasets and achieves state-of-the-art performance on three of them. |
Inducing Target-Specific Latent Structures for Aspect Sentiment Classification (2020.emnlp-main)
Copied to clipboard
| Challenge: | Aspect-level sentiment analysis aims to classify the sentiment polarity of an aspect or a target in a comment . graph convolutional networks can be used to classifice aspect terms in syllables . |
| Approach: | They propose to combine word dependency graphs and latent graphs to create latent models . they propose to model the interaction between the aspect and its surrounding contexts . |
| Outcome: | The proposed model can complement syntactic features with latent semantic dependencies. |