Cross-Domain Text Sentiment Analysis Based on CNN_FT Method
Transfer learning is one of the popular methods for solving the problem that the models built on the source domain cannot be directly applied to the target domain in the cross-domain sentiment classification. This paper proposes a transfer learning method based on the multi-layer convolutional neura...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-05-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/10/5/162 |
id |
doaj-a129720dc5ec447e904175f9efaca0da |
---|---|
record_format |
Article |
spelling |
doaj-a129720dc5ec447e904175f9efaca0da2020-11-25T01:36:39ZengMDPI AGInformation2078-24892019-05-0110516210.3390/info10050162info10050162Cross-Domain Text Sentiment Analysis Based on CNN_FT MethodJiana Meng0Yingchun Long1Yuhai Yu2Dandan Zhao3Shuang Liu4School of Computer Science & Engineering, Dalian Minzu University, Dalian 116600, ChinaSchool of Computer Science & Engineering, Dalian Minzu University, Dalian 116600, ChinaSchool of Computer Science & Engineering, Dalian Minzu University, Dalian 116600, ChinaSchool of Computer Science & Engineering, Dalian Minzu University, Dalian 116600, ChinaSchool of Computer Science & Engineering, Dalian Minzu University, Dalian 116600, ChinaTransfer learning is one of the popular methods for solving the problem that the models built on the source domain cannot be directly applied to the target domain in the cross-domain sentiment classification. This paper proposes a transfer learning method based on the multi-layer convolutional neural network (CNN). Interestingly, we construct a convolutional neural network model to extract features from the source domain and share the weights in the convolutional layer and the pooling layer between the source and target domain samples. Next, we fine-tune the weights in the last layer, named the fully connected layer, and transfer the models from the source domain to the target domain. Comparing with the classical transfer learning methods, the method proposed in this paper does not need to retrain the network for the target domain. The experimental evaluation of the cross-domain data set shows that the proposed method achieves a relatively good performance.https://www.mdpi.com/2078-2489/10/5/162cross-domainsentiment classificationtransfer learningconvolutional neural networkword2vec |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jiana Meng Yingchun Long Yuhai Yu Dandan Zhao Shuang Liu |
spellingShingle |
Jiana Meng Yingchun Long Yuhai Yu Dandan Zhao Shuang Liu Cross-Domain Text Sentiment Analysis Based on CNN_FT Method Information cross-domain sentiment classification transfer learning convolutional neural network word2vec |
author_facet |
Jiana Meng Yingchun Long Yuhai Yu Dandan Zhao Shuang Liu |
author_sort |
Jiana Meng |
title |
Cross-Domain Text Sentiment Analysis Based on CNN_FT Method |
title_short |
Cross-Domain Text Sentiment Analysis Based on CNN_FT Method |
title_full |
Cross-Domain Text Sentiment Analysis Based on CNN_FT Method |
title_fullStr |
Cross-Domain Text Sentiment Analysis Based on CNN_FT Method |
title_full_unstemmed |
Cross-Domain Text Sentiment Analysis Based on CNN_FT Method |
title_sort |
cross-domain text sentiment analysis based on cnn_ft method |
publisher |
MDPI AG |
series |
Information |
issn |
2078-2489 |
publishDate |
2019-05-01 |
description |
Transfer learning is one of the popular methods for solving the problem that the models built on the source domain cannot be directly applied to the target domain in the cross-domain sentiment classification. This paper proposes a transfer learning method based on the multi-layer convolutional neural network (CNN). Interestingly, we construct a convolutional neural network model to extract features from the source domain and share the weights in the convolutional layer and the pooling layer between the source and target domain samples. Next, we fine-tune the weights in the last layer, named the fully connected layer, and transfer the models from the source domain to the target domain. Comparing with the classical transfer learning methods, the method proposed in this paper does not need to retrain the network for the target domain. The experimental evaluation of the cross-domain data set shows that the proposed method achieves a relatively good performance. |
topic |
cross-domain sentiment classification transfer learning convolutional neural network word2vec |
url |
https://www.mdpi.com/2078-2489/10/5/162 |
work_keys_str_mv |
AT jianameng crossdomaintextsentimentanalysisbasedoncnnftmethod AT yingchunlong crossdomaintextsentimentanalysisbasedoncnnftmethod AT yuhaiyu crossdomaintextsentimentanalysisbasedoncnnftmethod AT dandanzhao crossdomaintextsentimentanalysisbasedoncnnftmethod AT shuangliu crossdomaintextsentimentanalysisbasedoncnnftmethod |
_version_ |
1725061708080218112 |