學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 Facebook上的情緒分析：台灣COVID-19期間疫苗貼文的研究
Sentiment Analysis on Facebook: A Case Study of Vaccine Posts during the COVID-19 in Taiwan作者 周廷威
Jhou, Ting-Wei貢獻者 邱淑怡
Chiu, Shu-I
周廷威
Jhou, Ting-Wei關鍵詞 COVID-19
社群媒體
疫苗
情緒分析
機器學習
COVID-19
Social media
Vaccine
Sentiment analysis
Machine learning日期 2024 上傳時間 4-九月-2024 15:02:23 (UTC+8) 摘要 本研究主要探討社群媒體之疫苗貼文的情緒分析，在疫情期間透過社群媒體接收疫情資訊日漸頻繁，快速獲得疫情新資訊的同時，也十分容易受貼文內容影響情緒，在許多人因疫情導致憂鬱、恐慌和害怕的時候，為了避免他們在收集解決狀況的資訊時心靈再次受到傷害，應該事先讓用戶可以知道他們瀏覽的貼文可能會帶給他們什麼樣的情緒。 本文將文本情感分為貼文作者和貼文讀者兩個方面的情感來進行分析： 作者方面視作貼文自身的情感，用雙向長短期記憶模型來學習社群媒體貼文的情感極性，BERT提取文本特徵後，透過Bi-LSTM模型進行訓練，隱藏層包括自注意力層、循環層和全連接層，模型結果呈現正向文本的預測非常的好，中立和負向的文本預測則是普通，會有預測錯誤的情形。 讀者方面使用Facebook reactions來計算對文本的好感程度，Facebook reactions本身即帶有情感定義，相關研究中也有被使用來分析讀者對貼本類別的喜好，但是疫情期間Facebook reactions新增了一個情感，因此計算正向和負向的公式已不適用，國外與台灣對於Facebook的使用習慣也有不同，因此要使用這個資料來進行情感分析需要重新修改情感定義和計算公式，調整過的公式在實驗得到的情感指數與疫情期間台灣對各廠牌疫苗的看法相似。
This study explores emotional analysis of vaccine-related social media posts during the pandemic. Excessive exposure to epidemic information on social media has made individuals susceptible to emotional influence. Amid feelings of depression, panic, and fear, it's vital to caution users about potential emotional impacts before viewing posts to prevent further distress while seeking solutions. The study assesses emotions from post authors' and readers' perspectives. A Bi-LSTM model with BERT for text feature extraction is used to determine emotional polarity. For reader analysis, Facebook reactions gauge post favorability. However, the pandemic's addition of new emotions in reactions challenges existing sentiment calculation formulas. Adjusted formulas in experiments resulted in sentiment indices akin to Taiwan's perceptions of vaccine brands during the pandemic.參考文獻 [1] Shu-I Chiu and Kuo-Wei Hsu. Information diffusion on facebook: A case study of the sunflower student movement in taiwan. In Proceedings of the 11th international conference on ubiquitous information management and communication, pages 1–8, 2017. [2] Eduardo C Costa, Alex B Vieira, Klaus Wehmuth, Artur Ziviani, and Ana Paula Couto Da Silva. Time centrality in dynamic complex networks. Advances in Complex Systems, 18(07n08):1550023, 2015. [3] Duduzile Ndwandwe and Charles S Wiysonge. Covid-19 vaccines. Current opinion in immunology, 71:111–116, 2021. [4] Rodrigo Sandoval-Almazan and David Valle-Cruz. Sentiment analysis of facebook users reacting to political campaign posts. Digital Government: Research and Prac tice, 1(2):1–13, 2020. [5] Serpil Aslan. A novel tcnn–bi-lstm deep learning model for predicting sentiments of tweets about covid-19 vaccines. Concurrency and Computation: Practice and Experience, 34(28):e7387, 2022. [6] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018. [7] MikeSchuster and Kuldip K Paliwal. Bidirectional recurrent neural networks. IEEE transactions on Signal Processing, 45(11):2673–2681, 1997. [8] Yahui Chen. Convolutional neural network for sentence classification. Master’s thesis, University of Waterloo, 2015. [9] Nitesh V Chawla, Kevin W Bowyer, Lawrence O Hall, and W Philip Kegelmeyer. Smote: syntheticminorityover-samplingtechnique. Journalofartificialintelligence research, 16:321–357, 2002. [10] AshishVaswani,NoamShazeer,NikiParmar,JakobUszkoreit,LlionJones,AidanN Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. [11] Chong Oh and Savan Kumar. How trump won: the role of social media sentiment in political elections. 2017. [12] Abhishek Singh, Eduardo Blanco, and Wei Jin. Incorporating emoji descriptions improves tweet classification. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Lan guage Technologies, Volume 1 (Long and Short Papers), pages 2096–2101, 2019. [13] Cole Freeman, Hamed Alhoori, and Murtuza Shahzad. Measuring the diversity of facebook reactions to research. Proceedings of the ACM on Human-Computer In teraction, 4(GROUP):1–17, 2020. [14] Li-Chen Cheng and Song-Lin Tsai. Deep learning for automated sentiment analysis of social media. In Proceedings of the 2019 IEEE/ACM international conference on advances in social networks analysis and mining, pages 1001–1004, 2019. [15] Mayur Wankhade, Annavarapu Chandra Sekhara Rao, and Chaitanya Kulkarni. A survey on sentiment analysis methods, applications, and challenges. Artificial Intel ligence Review, 55(7):5731–5780, 2022. [16] Song Xie, Jingjing Cao, Zhou Wu, Kai Liu, Xiaohui Tao, and Haoran Xie. Sen timent analysis of chinese e-commerce reviews based on bert. In 2020 IEEE 18th International Conference on Industrial Informatics (INDIN), volume 1, pages 713 718. IEEE, 2020. [17] Teo Susnjak. Applying bert and chatgpt for sentiment analysis of lyme disease in scientific literature. In Borrelia burgdorferi: Methods and Protocols, pages 173 183. Springer, 2024. [18] Xinying Chen, Peimin Cong, and Shuo Lv. A long-text classification method of chinese news based on bert and cnn. IEEE Access, 10:34046–34057, 2022. [19] Tianyi Wang, Ke Lu, Kam Pui Chow, and Qing Zhu. Covid-19 sensing: negative sentiment analysis on social media in china via bert model. Ieee Access, 8:138162 138169, 2020. [20] 徐秀and刘德喜. 基于上下文和位置交互协同注意力的文本情绪原因识别. 中 文信息学报,36(2):142–151, 2022. [21] 谭金源,刁宇峰,祁瑞华,and林鸿飞. 基于bert-pgn模型的中文新闻文本自动 摘要生成. 计算机应用,41(1):127–132,2021. 描述 碩士
國立政治大學
資訊科學系
111753223資料來源 http://thesis.lib.nccu.edu.tw/record/#G0111753223 資料類型 thesis dc.contributor.advisor 邱淑怡 zh_TW dc.contributor.advisor Chiu, Shu-I en_US dc.contributor.author (作者) 周廷威 zh_TW dc.contributor.author (作者) Jhou, Ting-Wei en_US dc.creator (作者) 周廷威 zh_TW dc.creator (作者) Jhou, Ting-Wei en_US dc.date (日期) 2024 en_US dc.date.accessioned 4-九月-2024 15:02:23 (UTC+8) - dc.date.available 4-九月-2024 15:02:23 (UTC+8) - dc.date.issued (上傳時間) 4-九月-2024 15:02:23 (UTC+8) - dc.identifier (其他 識別碼) G0111753223 en_US dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/153392 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊科學系 zh_TW dc.description (描述) 111753223 zh_TW dc.description.abstract (摘要) 本研究主要探討社群媒體之疫苗貼文的情緒分析，在疫情期間透過社群媒體接收疫情資訊日漸頻繁，快速獲得疫情新資訊的同時，也十分容易受貼文內容影響情緒，在許多人因疫情導致憂鬱、恐慌和害怕的時候，為了避免他們在收集解決狀況的資訊時心靈再次受到傷害，應該事先讓用戶可以知道他們瀏覽的貼文可能會帶給他們什麼樣的情緒。 本文將文本情感分為貼文作者和貼文讀者兩個方面的情感來進行分析： 作者方面視作貼文自身的情感，用雙向長短期記憶模型來學習社群媒體貼文的情感極性，BERT提取文本特徵後，透過Bi-LSTM模型進行訓練，隱藏層包括自注意力層、循環層和全連接層，模型結果呈現正向文本的預測非常的好，中立和負向的文本預測則是普通，會有預測錯誤的情形。 讀者方面使用Facebook reactions來計算對文本的好感程度，Facebook reactions本身即帶有情感定義，相關研究中也有被使用來分析讀者對貼本類別的喜好，但是疫情期間Facebook reactions新增了一個情感，因此計算正向和負向的公式已不適用，國外與台灣對於Facebook的使用習慣也有不同，因此要使用這個資料來進行情感分析需要重新修改情感定義和計算公式，調整過的公式在實驗得到的情感指數與疫情期間台灣對各廠牌疫苗的看法相似。 zh_TW dc.description.abstract (摘要) This study explores emotional analysis of vaccine-related social media posts during the pandemic. Excessive exposure to epidemic information on social media has made individuals susceptible to emotional influence. Amid feelings of depression, panic, and fear, it's vital to caution users about potential emotional impacts before viewing posts to prevent further distress while seeking solutions. The study assesses emotions from post authors' and readers' perspectives. A Bi-LSTM model with BERT for text feature extraction is used to determine emotional polarity. For reader analysis, Facebook reactions gauge post favorability. However, the pandemic's addition of new emotions in reactions challenges existing sentiment calculation formulas. Adjusted formulas in experiments resulted in sentiment indices akin to Taiwan's perceptions of vaccine brands during the pandemic. en_US dc.description.tableofcontents 1 Introduction 1 2 Related Work 8 2.1 BERT 8 2.2 Bi-LSTM 8 2.3 TCNN-Bi-LSTM 10 2.4 SMOTE 11 2.5 Attention Mechanism 11 2.6 The sentiment of Facebook’s reactions 12 2.7 Sentiment Index 13 3 Method 14 3.1 Data Collection 14 3.2 Labeling 16 3.3 Data Preprocessing 16 3.4 Model Construction 18 3.5 Defining the sentiment of Facebook reactions 20 3.6 Dissemination Level of Post Comments and Shares 23 4 Experiments 25 4.1 Sentiment model 25 4.2 Sentiment Index 32 5 Conclusions 40 Reference 45 zh_TW dc.format.extent 1776474 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0111753223 en_US dc.subject (關鍵詞) COVID-19 zh_TW dc.subject (關鍵詞) 社群媒體 zh_TW dc.subject (關鍵詞) 疫苗 zh_TW dc.subject (關鍵詞) 情緒分析 zh_TW dc.subject (關鍵詞) 機器學習 zh_TW dc.subject (關鍵詞) COVID-19 en_US dc.subject (關鍵詞) Social media en_US dc.subject (關鍵詞) Vaccine en_US dc.subject (關鍵詞) Sentiment analysis en_US dc.subject (關鍵詞) Machine learning en_US dc.title (題名) Facebook上的情緒分析：台灣COVID-19期間疫苗貼文的研究 zh_TW dc.title (題名) Sentiment Analysis on Facebook: A Case Study of Vaccine Posts during the COVID-19 in Taiwan en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1] Shu-I Chiu and Kuo-Wei Hsu. Information diffusion on facebook: A case study of the sunflower student movement in taiwan. In Proceedings of the 11th international conference on ubiquitous information management and communication, pages 1–8, 2017. [2] Eduardo C Costa, Alex B Vieira, Klaus Wehmuth, Artur Ziviani, and Ana Paula Couto Da Silva. Time centrality in dynamic complex networks. Advances in Complex Systems, 18(07n08):1550023, 2015. [3] Duduzile Ndwandwe and Charles S Wiysonge. Covid-19 vaccines. Current opinion in immunology, 71:111–116, 2021. [4] Rodrigo Sandoval-Almazan and David Valle-Cruz. Sentiment analysis of facebook users reacting to political campaign posts. Digital Government: Research and Prac tice, 1(2):1–13, 2020. [5] Serpil Aslan. A novel tcnn–bi-lstm deep learning model for predicting sentiments of tweets about covid-19 vaccines. Concurrency and Computation: Practice and Experience, 34(28):e7387, 2022. [6] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018. [7] MikeSchuster and Kuldip K Paliwal. Bidirectional recurrent neural networks. IEEE transactions on Signal Processing, 45(11):2673–2681, 1997. [8] Yahui Chen. Convolutional neural network for sentence classification. Master’s thesis, University of Waterloo, 2015. [9] Nitesh V Chawla, Kevin W Bowyer, Lawrence O Hall, and W Philip Kegelmeyer. Smote: syntheticminorityover-samplingtechnique. Journalofartificialintelligence research, 16:321–357, 2002. [10] AshishVaswani,NoamShazeer,NikiParmar,JakobUszkoreit,LlionJones,AidanN Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. [11] Chong Oh and Savan Kumar. How trump won: the role of social media sentiment in political elections. 2017. [12] Abhishek Singh, Eduardo Blanco, and Wei Jin. Incorporating emoji descriptions improves tweet classification. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Lan guage Technologies, Volume 1 (Long and Short Papers), pages 2096–2101, 2019. [13] Cole Freeman, Hamed Alhoori, and Murtuza Shahzad. Measuring the diversity of facebook reactions to research. Proceedings of the ACM on Human-Computer In teraction, 4(GROUP):1–17, 2020. [14] Li-Chen Cheng and Song-Lin Tsai. Deep learning for automated sentiment analysis of social media. In Proceedings of the 2019 IEEE/ACM international conference on advances in social networks analysis and mining, pages 1001–1004, 2019. [15] Mayur Wankhade, Annavarapu Chandra Sekhara Rao, and Chaitanya Kulkarni. A survey on sentiment analysis methods, applications, and challenges. Artificial Intel ligence Review, 55(7):5731–5780, 2022. [16] Song Xie, Jingjing Cao, Zhou Wu, Kai Liu, Xiaohui Tao, and Haoran Xie. Sen timent analysis of chinese e-commerce reviews based on bert. In 2020 IEEE 18th International Conference on Industrial Informatics (INDIN), volume 1, pages 713 718. IEEE, 2020. [17] Teo Susnjak. Applying bert and chatgpt for sentiment analysis of lyme disease in scientific literature. In Borrelia burgdorferi: Methods and Protocols, pages 173 183. Springer, 2024. [18] Xinying Chen, Peimin Cong, and Shuo Lv. A long-text classification method of chinese news based on bert and cnn. IEEE Access, 10:34046–34057, 2022. [19] Tianyi Wang, Ke Lu, Kam Pui Chow, and Qing Zhu. Covid-19 sensing: negative sentiment analysis on social media in china via bert model. Ieee Access, 8:138162 138169, 2020. [20] 徐秀and刘德喜. 基于上下文和位置交互协同注意力的文本情绪原因识别. 中 文信息学报,36(2):142–151, 2022. [21] 谭金源,刁宇峰,祁瑞华,and林鸿飞. 基于bert-pgn模型的中文新闻文本自动 摘要生成. 计算机应用,41(1):127–132,2021. zh_TW
