臺灣博碩士論文加值系統

在新型冠狀病毒的流行之下，全世界的人都在公開的社交媒體上分享他們的意見，然而不齊全的資訊將Covid-19 訊息四處散播，導致多數人對於新型冠狀病毒的蔓延歸咎於亞洲人。社交媒體是種族歧視觀點的主要來源，應該受到監督並適時發現種族歧視之言論。
本研究目標在推特社交媒體上檢測到包含種族歧視的推文，並對其數據進行情緒分析。蒐集的數據集包括兩個情緒類別標籤的分類：正面情緒和負面情緒。然而數據集中包含顯著優勢的正面情緒類別，針對這個問題，本實驗使用了隨機過採樣(random oversampling)、隨機欠採樣(random under sampling)和合成少數過採樣(synthetic minority oversampling)三種抽樣方法。
本研究中使用機器學習和深度學習方法，且提出了具有不同內核但相同架構的卷積層CNN模型，另外，實驗還提出了具有不同池化大小的 CNN-LSTM 方法。實驗的結果表明，隨機森林在整體機器學習和抽樣方法方面取得了最好的分類分數，然而，深度學習方法的最佳效能來自內核為5的CNN方法，可產生90%的準確率。綜合以上所述，使用正確的內核和池化大小設置會對結果產生特定的影響。

In this pandemic condition of Corona, users around the world are sharing their opinions on social media. However, the spreading of Covid-19 led to a situation that most of the opinions are to blame Asians in this regard. Social media, being the leading source of racist opinions, should be monitored and racist remarks should be detected. This study aims to do a sentiment analysis of the data detected from tweets containing racist texts. The datasets include texts labeled according to two sentiment classes: positive and negative. However, the dataset contains a significant predominance of the positive class. In dealing with this issue, three resampling techniques are used which are random oversampling, random sampling, and synthetic minority oversampling. Techniques performed in this study are machine learning and deep learning. A convolutional layer of the CNN method is proposed with different kernels but the same architecture. In addition, the CNN-LSTM method is proposed with different pool sizes. Experimental results reveal that Random Forest achieves the best classification scores for overall machine learning and sampling methods. However, the best performance of the deep learning method comes from the CNN method with kernel 5 producing 90% accuracy. In conclusion, the right kernel and pool size setting leads to a specific effect on the result.

Abstract...i
摘要...ii
Acknowledgement...iii
Table of contents...iv
List of Tables...vi
List of Figure...vii
Chapter 1 Introduction...1
1.1 Background...1
1.2 Related Work...3
1.3 Research Purpose...5
1.4 Research Structure...6
Chapter 2 Literature Review...7
2.1 Sentiment Analysis...7
2.1.1 Introduction of Sentiment Analysis...7
2.1.2 Sentiment Analysis in COVID-19 issue...7
2.2 Natural Language Processing...8
2.2.1 Data Preprocessing...9
2.3 Machine Learning...10
2.3.1 Naïve Bayes...10
2.3.2 Decision Tree...11
2.3.3 Logistic Regression...12
2.3.4 Random Forest...12
2.4 Deep Learning Method...13
2.4.1 Long Short-Term Memory (LSTM)...14
2.4.2 Convolutional Neural Network (CNN)...14
2.4.3 Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM)...16
2.5 Convolutional Layer...16
2.6 Performance metric of sentiment classification...17
2.5 Sampling for imbalance data...18
2.5.1 Random Over Sampling (ROS)...19
2.5.2 Random Under Sampling (RUS)...19
2.5.3 Synthetic Minority Oversampling Technique (SMOTE)...20
Chapter 3 Research Methodology...21
3.1 Experimental Design...21
3.2 Data extraction through Twitter API...23
3.3 Data collection...24
3.4 Pre-processing of Twitter Data...24
3.5 Machine Learning Method...26
3.6 Deep Learning Method...28
3.6.1 Experimental design in LSTM...30
3.6.2 Experimental design in kernel (CNN)...31
3.6.3 Pooling size experiment in CNN-LSTM Method...34
Chapter 4 Result and Analysis...38
4.1 Tweet collected from twitter and online dataset...38
4.2 Each tweet pre-processing result...41
4.3 Machine Learning Result...46
4.4 Deep Learning Result...48
4.6 Kernel and Pooling Performance...50
Chapter 5 Conclusions and Future Work...53
5.1 Conclusions...53
5.2 Future Work...54
References...55
Appendix I – Machine Learning Result...61
Appendix II– Deep Learning Result...69
Extended Abstract...79

[1] A., P., &Priya, K. S. (2020). Sentiment Classification of Tweets in Twitter using CNN and Dropouts in RNN. International Journal of Computer Applications, 175(33), 1–5. https://doi.org/10.5120/ijca2020920681
[2] Aarti, M., &Patil, A. (2014). Sentiment Analysis for Product Reviews. International Journal of Advanced Research in Computer Science, 5(5), 202–204. https://doi.org/10.21917/ijsc.2019.0266
[3] Al-Hassan, A., &Al-Dossari, H. (2019). Detection of Hate Speech in Social Networks: a Survey on Multilingual Corpus. 83–100. https://doi.org/10.5121/csit.2019.90208
[4] Alotaibi, A., &Abul Hasanat, M. H. (2020). Racism Detection in Twitter Using Deep Learning and Text Mining Techniques for the Arabic Language. Proceedings - 2020 1st International Conference of Smart Systems and Emerging Technologies, SMART-TECH 2020, 161–164. https://doi.org/10.1109/SMART-TECH49988.2020.00047
[5] Benítez-Andrades, J. A., González-Jiménez, Á., López-Brea, Á., Aveleira-Mata, J., Alija-Pérez, J. M., &García-Ordás, M. T. (2022). Detecting racism and xenophobia using deep learning models on Twitter data: CNN, LSTM and BERT. PeerJ Computer Science, 8, 1–24. https://doi.org/10.7717/PEERJ-CS.906
[6] Bliuc, A. M., Faulkner, N., Jakubowicz, A., &McGarty, C. (2018). Online networks of racial hate: A systematic review of 10 years of research on cyber-racism. Computers in Human Behavior, 87, 75–86. https://doi.org/10.1016/j.chb.2018.05.026
[7] Çelik, Ö. (2018). A Research on Machine Learning Methods and Its Applications. Journal of Educational Technology and Online Learning, September 2018. https://doi.org/10.31681/jetol.457046
[8] Dr. S. Vijayarani, Ms. J. Ilamathi, M. N. (2015). Preprocessing Techniques for Text Mining Preprocessing Techniques for Text Mining. International Journal of Computer Science & Communication Networks, 5(October 2014), 7–16.
[9] Fang, H., Lu, C., Hong, F., Jiang, W., &Wang, T. (2021). Convolutional Neural Network for Heartbeat Classification. 2021 15th IEEE International Conference on Electronic Measurement and Instruments, ICEMI 2021, 253–258. https://doi.org/10.1109/ICEMI52946.2021.9679581
[10] Forrester, S., Jacobs, D., Zmora, R., Schreiner, P., Roger, V., &Kiefe, C. I. (2019). Racial differences in weathering and its associations with psychosocial stress: The CARDIA study. SSM - Population Health, 7(August 2018), 100319. https://doi.org/10.1016/j.ssmph.2018.11.003
[11] Garcia, K., &Berton, L. (2021). Topic detection and sentiment analysis in Twitter content related to COVID-19 from Brazil and the USA. Applied Soft Computing, 101, 107057. https://doi.org/10.1016/j.asoc.2020.107057
[12] Hassan, A., &Mahmood, A. (2018). Convolutional Recurrent Deep Learning Model for Sentence Classification. IEEE Access, 6(April), 13949–13957. https://doi.org/10.1109/ACCESS.2018.2814818
[13] He, B., Ziems, C., Soni, S., Ramakrishnan, N., Yang, Di., &Kumar, S. (2021). Racism is a virus: Anti-asian hate and counterspeech in social media during the COVID-19 crisis@gatech.edu. Proceedings of the 2021 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, ASONAM 2021, 90–94. https://doi.org/10.1145/3487351.3488324
[14] Jagdale, R. S., Shirsat, V. S., &Deshmukh, S. N. (2019). Sentiment analysis on product reviews using machine learning techniques. In Advances in Intelligent Systems and Computing (Vol. 768, Issue January). Springer Singapore. https://doi.org/10.1007/978-981-13-0617-4_61
[15] Kaiser, K. R., Kaiser, D. M., Kaiser, R. M., &Rackham, A. M. (2018). Using Social Media to Understand and Guide the Treatment of Racist Ideology. Global Journal of Guidance and Counseling in Schools: Current Perspectives, 8(1), 38–49. https://doi.org/10.18844/gjgc.v8i1.3579
[16] Khan, W., Daud, A., Nasir, J. A., &Amjad, T. (2016). A survey on the state-of-the-art machine learning models in the context of NLP. Kuwait Journal of Science, 43(4), 95–113.
[17] Khurana, D., Koli, A., Khatter, K., &Singh, S. (2022). Natural language processing: state of the art, current trends and challenges. Multimedia Tools and Applications, April 2018. https://doi.org/10.1007/s11042-022-13428-4
[18] Kim, H., &Jeong, Y. S. (2019). Sentiment classification using Convolutional Neural Networks. Applied Sciences (Switzerland), 9(11), 1–14. https://doi.org/10.3390/app9112347
[19] Kumari, K., Singh, J. P., Dwivedi, Y. K., &Rana, N. P. (2021). Bilingual Cyber-aggression detection on social media using LSTM autoencoder. Soft Computing, 25(14), 8999–9012. https://doi.org/10.1007/s00500-021-05817-y
[20] Lakshmanarao, A., Srisaila, A., &Kiran, T. S. R. (2022). Twitter Sentiment Classification with Deep Learning LSTM for Airline Tweets. 8th International Conference on Advanced Computing and Communication Systems, ICACCS 2022, 520–524. https://doi.org/10.1109/ICACCS54159.2022.9785208
[21] Lee, E., Rustam, F., Washington, P. B., Barakaz, F.El, Aljedaani, W., &Ashraf, I. (2022). Racism Detection by Analyzing Differential Opinions Through Sentiment Analysis of Tweets Using Stacked Ensemble GCR-NN Model. IEEE Access, 10(January), 9717–9728. https://doi.org/10.1109/ACCESS.2022.3144266
[22] Liu, B., Blasch, E., Chen, Y., Shen, D., &Chen, G. (2013). Scalable sentiment classification for Big Data analysis using Naïve Bayes Classifier. Proceedings - 2013 IEEE International Conference on Big Data, Big Data 2013, 99–104. https://doi.org/10.1109/BigData.2013.6691740
[23] Najafabadi, M. M., Villanustre, F., Khoshgoftaar, T. M., Seliya, N., Wald, R., &Muharemagic, E. (2015). Deep learning applications and challenges in big data analytics. Journal of Big Data, 2(1), 1–21. https://doi.org/10.1186/s40537-014-0007-7
[24] Nath, B., Shri, K., Vaishno, M., Wankhade, M., Chandra, A., Rao, S., Dara, S., &Kaushik, B. (2017). A Sentiment Analysis of Food Review using Logistic Regression. 2(7), 251–260. https://www.researchgate.net/publication/334654833
[25] Plaza-Del-Arco, F. M., Molina-González, M. D., Ureña-López, L. A., &Martín-Valdivia, M. T. (2020). Detecting Misogyny and Xenophobia in Spanish Tweets Using Language Technologies. ACM Transactions on Internet Technology, 20(2). https://doi.org/10.1145/3369869
[26] Poornima, A., &Priya, K. S. (2020). A Comparative Sentiment Analysis of Sentence Embedding Using Machine Learning Techniques. 2020 6th International Conference on Advanced Computing and Communication Systems, ICACCS 2020, 493–496. https://doi.org/10.1109/ICACCS48705.2020.9074312
[27] Putri, T. T. A., Sriadhi, S., Sari, R. D., Rahmadani, R., &Hutahaean, H. D. (2020). A comparison of classification algorithms for hate speech detection. IOP Conference Series: Materials Science and Engineering, 830(3). https://doi.org/10.1088/1757-899X/830/3/032006
[28] Saha, B. N., Senapati, A., &Mahajan, A. (2020). LSTM based Deep RNN Architecture for Election Sentiment Analysis from Bengali Newspaper. 2020 International Conference on Computational Performance Evaluation, ComPE 2020, 564–569. https://doi.org/10.1109/ComPE49325.2020.9200062
[29] Samat, N. A., Salleh, M. N. M., &Ali, H. (2020). The Comparison of Pooling Functions in Convolutional Neural Network for Sentiment Analysis Task. Advances in Intelligent Systems and Computing, 978 AISC(January), 202–210. https://doi.org/10.1007/978-3-030-36056-6_20
[29] Samuel, J., Ali, G. G. M. N., Rahman, M. M., Esawi, E., &Samuel, Y. (2020). COVID-19 public sentiment insights and machine learning for tweets classification. Information (Switzerland), 11(6), 1–22. https://doi.org/10.3390/info11060314
[30] Shu, B., Ren, F., &Bao, Y. (2018). Investigating Lstm with k-Max Pooling for Text Classification. Proceedings - 11th International Conference on Intelligent Computation Technology and Automation, ICICTA 2018, 31–34. https://doi.org/10.1109/ICICTA.2018.00015
[31] Sosa, P. M. (2017). Twitter Sentiment Analysis using combined LSTM-CNN Models. Eprint Arxiv, 1–9.
[32] Sumathy, P., &Muthukumari, S. M. (2018). Sentiment Analysis of Twitter Data Using Multi Class Semantic Approach. International Journal of Scientific Research in Computer Science, Engineering and Information Technology © 2018 IJSRCSEIT |, 3(10), 2456–3307. www.amazon.com
[33] Koehrsen, W. (2017). Retrieved from https://williamkoehrsen.medium.com/random-forest-simple-explanation-377895a60d2d
[34] Kim, Y., Jernite, Y., Sontag, D., & Rush, A. M. (2016, March). Character-aware neural language models. In Thirtieth AAAI conference on artificial intelligence. https://doi.org/10.48550/arXiv.1508.06615
[35] Github. Retrieved from https://cs231n.github.io/convolutional-networks/#pool
[36] Olah, C. (2015) Retrieved from http://colah.github.io/posts/2015-08-Understanding-LSTMs/
[37] Analytics Vidhya. Retrieved from https://medium.com/analytics-vidhya/undersampling-and-oversampling-an-old-and-a-new-approach-4f984a0e8392
[38] Cupoy. (2021) Retrieved from https://www.cupoy.com/qa/club/ai_tw/0000016D6BA22D97000000016375706F795F72656C656173654B5741535354434C5542/0000017BD8FC94950000000A6375706F795F72656C656173655155455354