社群媒體在過去十年中,已漸成為使用者表達和分享見解的主要工具。隨著社群媒
體的發展擴增,如何從平台上擷取新信息以促進、評估並反饋使用者交流的成果,
是現下各家角逐的新藍海;目前,社群計算是常用於解決此類任務的做法。社群運
算的其中一支做法,主要在識別使用者生成文章的內容,從中了解使用者的想法;
而分析使用者撰文中的責備語意,能有助於了解使用者的社會行為,對社群的認知
並且進一步規範社群的行為。

在本研究中,我們提出一項能夠偵測社群媒體中責備表達的新方法。我
們的模型能夠基於社群媒體「推特」上的使用者貼文擷取語意特徵,以辨識責備與
咎責歸因是否存在。且基於圖論製作的語意樣式,能夠使用非監督式的方式,萃取文句
中潛藏有咎責或歸因意涵的內容。我們的實驗結果顯示,儘管沒有使用標註資料或
詞義資料輔助訓練,我們提出的模型能夠有效率的從資料中,辨識出「自我歸咎」
與「歸咎他人」兩種類型的咎責歸因。

Social networking platforms have become major hubs for users to express their insights. The growth of such platforms has drawn a lot of attention to the need for facilitating and evaluating social behaviors and user interactions in an effort to obtain new information. Social computing is an area capable of solving such tasks computationally. A major element of social computing is identifying attributions of the blame on user-generated content. It is important to analyze attributions of blame because it helps to understand other people' social behaviors, regulates social conduct and cognitive aspects of community members. In this research, we present an innovative approach of identifying blame on social media. Model generates linguistic features based on the user-generated content of Twitter. These features are able to identify possible patterns of blame expressions without the need of having an annotated corpus. The experimental result shows that our model can efficiently generate syntactic patterns representing two types of blame: self-blame and blame-others; despite not using annotated data and lexicon-based approaches such are LIWC, POS, etc.

References
[1] KristineCanales,JoEllenPope,andCherieD.Maestas.Tweetingblameinafederalist system: Attributions for disaster response in social media following hurricane sandy*. Social Science Quarterly, 05 2019.
[2] Rich Caruana, Nikos Karampatziakis, and Ainur Yessenalina. An empirical evalua- tion of supervised learning in high dimensions. pages 96–103, 2008.
[3] Mario Cataldi, Luigi Di Caro, and Claudio Schifanella. Emerging topic detection on twitter based on temporal and social terms evaluation. page 4, 2010.
[4] Nicholas Diakopoulos, Mor Naaman, and Funda Kivran-Swaine. Diamonds in the rough: Social media visual analytics for journalistic inquiry. pages 115–122, 2010.
[5] AlecGo,RichaBhayani,andLeiHuang.Twittersentimentclassificationusingdistant supervision. CS224N Project Report, Stanford, 1(12):2009, 2009.
[6] ArthurGraesser,DanielleMcNamara,MaxLouwerse,andZhiqiangCai.Coh-metrix: Analysis of text on cohesion and language. Behavior research methods, instruments, computers : a journal of the Psychonomic Society, Inc, 36:193–202, 06 2004.
49
[7] Sandra Graham and Jaana Juvonen. Self-blame and peer victimization in middle school: an attributional analysis. Developmental psychology, 34(3):587, 1998.
[8] Ronnie Janoff-Bulman. Characterological versus behavioral self-blame: Inquiries into depression and rape. Journal of personality and social psychology, 37:1798– 809, 11 1979.
[9] Karin Kipper, Anna Korhonen, Neville Ryant, and Martha Palmer. Extending verbnet with novel verb classes. 01 2006.
[10] WW Leontief. The structure of the american economy, 1919–1929 harvard university press. Cambridge (new, enlarged edition, Oxford University Press, New York, 1951), 1941.
[11] Bertram Malle, Steve Guglielmo, and Andrew Monroe. A theory of blame. Psycho- logical Inquiry, 25:147–186, 04 2014.
[12] Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and David McClosky. The Stanford CoreNLP natural language processing toolkit. pages 55–60, 2014.
[13] Wenji Mao, Ansheng Ge, and Xiaochen Li. From causal scenarios to social causality: An attributional approach. IEEE Intelligent Systems, 26:48–57, 11 2011.
[14] Udochukwu Orizu and Yulan He. Detecting expressions of blame or praise in text. pages 4124–4129, May 2016.
[15] Mark D Pagel, Joseph Becker, and David B Coppel. Loss of control, self-blame,
50

and depression: An investigation of spouse caregivers of alzheimer’s disease patients. Journal of Abnormal Psychology, 94(2):169, 1985.
[16] Elvis Saravia, Hsien-Chi Toby Liu, Yen-Hao Huang, Junlin Wu, and Yi-Shin Chen. Carer: Contextualized affect representations for emotion recognition. pages 3687– 3697, 2018.
[17] John R Seeley. The net of reciprocal influence. a problem in treating sociometric data. Canadian Journal of Experimental Psychology, 3:234, 1949.
[18] Yla R. Tausczik and James Pennebaker. The psychological meaning of words: Liwc and computerized text analysis methods. Journal of Language and Social Psychology, 29(1):24–54, 3 2010.
[19] Leah C. Windsor, Nia Dowell, and Art Graesser. The language of autocrats: Leaders’ language in natural disaster crises. Risk, Hazards & Crisis in Public Policy, 5(4):446– 467, 2014.
[20] Zheng Xu, Hui Zhang, Vijayan Sugumaran, Kim-Kwang Raymond Choo, Lin Mei, and Yiwei Zhu. Participatory sensing-based semantic and spatial analysis of urban emergency events using mobile social media. EURASIP Journal on Wireless Com- munications and Networking, 2016(1):44, 2016.