社交媒體是用戶向世界展示自己的地方，揭示個人細節和對他們生活 的見解。
近年來，Instagram已成為一種流行的社交網絡應用程式，此允許 人們通過發布圖片來表達自己。
在本論文中，我們只對他們對人格測驗感興趣其僅利用他們的圖 片內容做分析。
人格測驗已被證明與許多類型的互動有關;
它已被證明有助於預測工作滿意度，專業和浪漫的關係成功，甚 至偏好不同的界面。
為了收集數據，我們使用不同的主題標籤和在此主題標籤上發布 的用戶的內容收集了內容。
我們收集了來自2,735名用戶的圖像和總共20,233份用戶，經過整 理過和再預先處理後，我們最終獲得了2,294張圖片和16,203張圖片。
i
在本論文中，我們提出了一種方法，通過該方法可以藉 由Instagram個人資料中的公開圖片內容而準確預測用戶的個性。
我們將描述收集的數據類型，收集方式，分析方法以及通過機器 學習預測人格特質的結果。
我們提出了一個框架，使用BigFive模型利用在Instagram個人資料 中的公開圖片內容來預測人格特質。
我們將這些結果與之前的工作進行了比較，以衡量結果的準確 性。

Social media is a place where users show themselves to the world, reveal- ing personal details and insights into their lives. In the recent years Insta- gram has become a popular social network application that allows people to express themselves through posting pictures. In this paper, we are inter- ested in personality predictions using only their pictures content. Person- ality has been shown to be relevant to many types of interactions; it has been shown to be useful in predicting job satisfaction, professional and ro- mantic relationship success, and even preference for different interfaces. To collect the data, we collected the content using different hasthtags and the content of the users posting on this hastagas. We collected images from 2,735 Instragram profiles and a total of 20,233 images that after cleaning and pre-processing we end up with 2,294 valid Instagram profiles and a to- tal of 16,203 images. In this paper, we present a method by which a user’s
iii
personality can be accurately predicted through the publicly available pic- ture content from their Instagram profiles. We will describe the type of data collected,how was collected, our methods of analysis, and the results of predicting personality traits through machine learning. We propose a framework to predict personality traits using BigFive model using publicly available picture content from their Instagram profiles. We compared these results to previous work to measure the accuracy of the results.

摘要 i
Abstract iii
Acknowledgement v
List of Tables ix
List of Figures xi
1 Introduction ...............1
2 Related Work ........... 3
2.1 Big Five personality traits and social networking . . . . . . . 3
2.2 Personality Prediction on Social Media. . . . . . . . . . . . 4

3 Methodology ......... 6
3.1 Data collection Methodology................. 6
3.2 Personality Traits Detection.................. 7
3.2.1 How It predicts personality.............. 8
3.2.2 What is the performance ............... 9
3.3 VGGNet............................ 10
3.4 Our Network ......................... 12

4 Experiments and Results ............16
4.0.1 Ground Truth......................... 16
4.0.2 Prior Work Personality Prediction . . . . . . . . . . 18
4.0.3 PersonalityPredictionModel. . . . . . . . . . . . . 19
4.0.4 How number of images affects performace . . . . . 23
4.0.5 Real Examples .................... 24

5 Discussion ................................33
6 Conclusion and Future Work ....37
6.1 Conclusion .......................... 37
6.2 FutureWorkandLimitation ................. 39

References..................41

7 Appendix..................46

[1] Bruce Ferwerda and Marko Tkalcic. You are what you post: What the content of instagram pictures tells about users’ personality. In The 23rd International on Intelligent User Interfaces, March 7-11, Tokyo, Japan, 2018.
[2] Shannon Greenwood, Andrew Perrin, and Maeve Duggan. Social me- dia update 2016. Pew Research Center, 11(2), 2016.
[3] Rong Hu and Pearl Pu. Enhancing collaborative filtering systems with personality information. In Proceedings of the fifth ACM conference on Recommender systems, pages 197–204. ACM, 2011.
[4] Denis Bratko, Ana Butkovic, and Michael Bosnjak. Twin study of im- pulsive buying and its overlap with personality. Journal of Individual Differences, 2013.
40
[5] Marko Tkalcˇicˇ, Bruce Ferwerda, David Hauger, and Markus Schedl. Personality correlates for digital concert program notes. In Interna- tional Conference on User Modeling, Adaptation, and Personalization, pages 364–369. Springer, 2015.
[6] Bruce Ferwerda and Marko Tkalcic. Predicting users’ personality from instagram pictures: Using visual and/or content features? In Proceed- ings of the 26th Conference on User Modeling, Adaptation and Per- sonalization, pages 157–161. ACM, 2018.
[7] Robert R McCrae and Oliver P John. An introduction to the five-factor model and its applications. Journal of personality, 60(2):175–215, 1992.
[8] Craig Ross, Emily S Orr, Mia Sisic, Jaime M Arseneault, Mary G Simmering, and R Robert Orr. Personality and motivations associated with facebook use. Computers in human behavior, 25(2):578–586, 2009.
[9] Yair Amichai-Hamburger, Hadar Kaplan, and Nira Dorpatcheon. Click to the past: The impact of extroversion by users of nostalgic websites
on the use of internet social services. Computers in Human Behavior, 24(5):1907–1912, 2008.
[10] Daniele Quercia, Michal Kosinski, David Stillwell, and Jon Crowcroft. Our twitter profiles, our selves: Predicting personality with twitter. In 2011 IEEE third international conference on privacy, security, risk and trust and 2011 IEEE third international conference on social comput- ing, pages 180–185. IEEE, 2011.
[11] Jennifer Golbeck, Cristina Robles, and Karen Turner. Predicting per- sonality with social media. In CHI’11 extended abstracts on human factors in computing systems, pages 253–262. ACM, 2011.
[12] Michal Kosinski, David Stillwell, and Thore Graepel. Private traits and attributes are predictable from digital records of human behavior. Pro- ceedings of the National Academy of Sciences, 110(15):5802–5805, 2013.
[13] Fabio Celli, Elia Bruni, and Bruno Lepri. Automatic personality and interaction style recognition from facebook profile pictures. In Pro- ceedings of the 22nd ACM international conference on Multimedia, pages 1101–1104. ACM, 2014.
[14] Bruce Ferwerda, Emily Yang, Markus Schedl, and Marko Tkalcic. Per- sonality traits predict music taxonomy preferences. In Proceedings of the 33rd Annual ACM Conference Extended Abstracts on Human Fac- tors in Computing Systems, pages 2241–2246. ACM, 2015.
[15] Bruce Ferwerda, Marko Tkalcic, and Markus Schedl. Personality traits and music genres: What do people prefer to listen to? In Proceedings of the 25th Conference on User Modeling, Adaptation and Personal- ization, pages 285–288. ACM, 2017.
[16] Pierre-Hadrien Arnoux, Anbang Xu, Neil Boyette, Jalal Mahmud, Rama Akkiraju, and Vibha Sinha. 25 tweets to know you: A new model to predict personality with social media. 2017.
[17] H Andrew Schwartz, Johannes C Eichstaedt, Margaret L Kern, Lukasz Dziurzynski, Stephanie M Ramones, Megha Agrawal, Achal Shah, Michal Kosinski, David Stillwell, Martin EP Seligman, et al. Per- sonality, gender, and age in the language of social media: The open- vocabulary approach. PloS one, 8(9):e73791, 2013.
[18] Barbara Plank and Dirk Hovy. Personality traits on twitter—or—how
to get 1,500 personality tests in a week. In Proceedings of the 6th Work- shop on Computational Approaches to Subjectivity, Sentiment and So- cial Media Analysis, pages 92–98, 2015.
[19] Karen Simonyan and Andrew Zisserman. Very deep convolu- tional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014.
[20] Kathie M Cole, Anna Gawlinski, Neil Steers, and Jenny Kotlerman. Animal-assisted therapy in patients hospitalized with heart failure. American Journal of critical care, 16(6):575–585, 2007.
[21] Margo A Halm. The healing power of the human-animal connection. American Journal of Critical Care, 17(4):373–376, 2008.