Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification
Aspect-Based (also known as aspect-level) Sentiment Classification (ABSC) aims at determining the sentimental tendency of a particular target in a sentence. With the successful application of the attention network in multiple fields, attention-based ABSC has aroused great interest. However, most of...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-03-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/10/6/2052 |
id |
doaj-e35aa212a103412eb61ef7ec69a9a18c |
---|---|
record_format |
Article |
spelling |
doaj-e35aa212a103412eb61ef7ec69a9a18c2020-11-25T02:01:59ZengMDPI AGApplied Sciences2076-34172020-03-01106205210.3390/app10062052app10062052Multiple Interactive Attention Networks for Aspect-Based Sentiment ClassificationDianyuan Zhang0Zhenfang Zhu1Qiang Lu2Hongli Pei3Wenqing Wu4Qiangqiang Guo5School of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaSchool of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaSchool of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaSchool of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaSchool of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaSchool of Information Science and Electrical Engineering, Shandong Jiao Tong University, Jinan 250357, ChinaAspect-Based (also known as aspect-level) Sentiment Classification (ABSC) aims at determining the sentimental tendency of a particular target in a sentence. With the successful application of the attention network in multiple fields, attention-based ABSC has aroused great interest. However, most of the previous methods are difficult to parallelize, insufficiently obtain, and fuse the interactive information. In this paper, we proposed a Multiple Interactive Attention Network (MIN). First, we used the Bidirectional Encoder Representations from Transformers (BERT) model to pre-process the data. Then, we used the partial transformer to obtain a hidden state in parallel. Finally, we took the target word and the context word as the core to obtain and fuse the interactive information. Experimental results on the different datasets showed that our model was much more effective.https://www.mdpi.com/2076-3417/10/6/2052pre-trained bertnatural language processingaspect-based sentiment classificationattention mechanism |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Dianyuan Zhang Zhenfang Zhu Qiang Lu Hongli Pei Wenqing Wu Qiangqiang Guo |
spellingShingle |
Dianyuan Zhang Zhenfang Zhu Qiang Lu Hongli Pei Wenqing Wu Qiangqiang Guo Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification Applied Sciences pre-trained bert natural language processing aspect-based sentiment classification attention mechanism |
author_facet |
Dianyuan Zhang Zhenfang Zhu Qiang Lu Hongli Pei Wenqing Wu Qiangqiang Guo |
author_sort |
Dianyuan Zhang |
title |
Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification |
title_short |
Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification |
title_full |
Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification |
title_fullStr |
Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification |
title_full_unstemmed |
Multiple Interactive Attention Networks for Aspect-Based Sentiment Classification |
title_sort |
multiple interactive attention networks for aspect-based sentiment classification |
publisher |
MDPI AG |
series |
Applied Sciences |
issn |
2076-3417 |
publishDate |
2020-03-01 |
description |
Aspect-Based (also known as aspect-level) Sentiment Classification (ABSC) aims at determining the sentimental tendency of a particular target in a sentence. With the successful application of the attention network in multiple fields, attention-based ABSC has aroused great interest. However, most of the previous methods are difficult to parallelize, insufficiently obtain, and fuse the interactive information. In this paper, we proposed a Multiple Interactive Attention Network (MIN). First, we used the Bidirectional Encoder Representations from Transformers (BERT) model to pre-process the data. Then, we used the partial transformer to obtain a hidden state in parallel. Finally, we took the target word and the context word as the core to obtain and fuse the interactive information. Experimental results on the different datasets showed that our model was much more effective. |
topic |
pre-trained bert natural language processing aspect-based sentiment classification attention mechanism |
url |
https://www.mdpi.com/2076-3417/10/6/2052 |
work_keys_str_mv |
AT dianyuanzhang multipleinteractiveattentionnetworksforaspectbasedsentimentclassification AT zhenfangzhu multipleinteractiveattentionnetworksforaspectbasedsentimentclassification AT qianglu multipleinteractiveattentionnetworksforaspectbasedsentimentclassification AT honglipei multipleinteractiveattentionnetworksforaspectbasedsentimentclassification AT wenqingwu multipleinteractiveattentionnetworksforaspectbasedsentimentclassification AT qiangqiangguo multipleinteractiveattentionnetworksforaspectbasedsentimentclassification |
_version_ |
1724954535603994624 |