Twitter 1,a) 1,b) Twitter Weibo Twitter Twitter 13 Twitter An Investigation and Analysis on the Relationship between Profile Images and User Behaviors on Twitter Tominaga Tomu 1,a) Hijikata Yoshinori 1,b) Abstract: In recent years, many researches on user profiling based on users usage objectives or link properties have arisen on micro-blog services such as Twitter or Weibo. These studies assume that users behaviors (e.g. posting, sharing) are influenced by their characteristics such as their demographic information, personality, or usage objectives. We believe that these characteristics of users also have an impact on what kind of images they set as their profile images on micro-blog services. This is because profile images are regarded as visual symbols of the owners. In this study, we firstly categorize profile images on Twitter into 13 types, then investigate the relationship between the types of profile images and users behaviors of those who belong to the type. We find several statistically significant differences of users behaviors across types of profile images. Keywords: Profile images, User behaviors, Twitter, User profiling 1. Twitter *1 Twitter 1 1-3, Machikane-yama, Toyonaka, Osaka, Japan a) tominaga@nishilab.sys.es.osaka-u.ac.jp b) hijikata@sys.es.osaka-u.ac.jp *1 https://twitter.com Twitter [1], [2], [5] [7], [9], [10], [11] Processing Society of Japan 272
[15] Twitter Twitter 13 13 4 5 2. *2 [4], [6], [13], [14], [16] [1], [2], [3], [8], [10], [11], [12] 3. 13 1 13 10 *2 SNS 1: (On) (Sp) (Hf) (As) (Dp) (Le) (Lo) (Ot) (Ch) (An) (Ob) (Sc) (Eg) Twitter 2: 300 Twitter Siegel 0.70 13 4. 13 Twitter 100 Twitter REST API Twitter Twitter 2 2014 2 23 3 25 N tweet N rt N rted N reply 5. 5.1 Twitter 1 Processing Society of Japan 273
R tw 0 10 20 30 40 50 60 70 R rt 0.0 0.1 0.2 0.3 0.4 0.5 0.6 1: 1 R tw Twitter REST API N AllT weet Days R tw R tw = N AllT weets / Days (1) 1 R tw 1 R tw Eg Ot R tw Steel-Dwass 3 R tw R tw R tw On Ot 6.55 p < 1.0 10 7 Ot 1 Lo R tw 2 5.2 Twitter 2: R rt R rt R rt = N rt / N tweet (2) R rt 2 R rt 1 R rt 1 R tw R rt 3 R rt 10 20 1 Eg Twitter 5.3 Twitter R rt R rted R rted = N rted / N tweet (3) Processing Society of Japan 274
3: t t n... n = 1 p < 0.05 n 2 p < 1.0 10 n R tw R rt R rted R reply Median Ot Lo Median Eg Median Lo Eg Median As Lo (On) 3.69 6.55 7 1.23 0.079 7.55 11 0.069 3.71 1 8.39 12 0.365 4.42 3 6.41 7 (Sp) 5.19 4.94 4 2.70 0.084 7.44 11 0.030 4.65 3 6.82 9 0.270 5.64 5 5.29 5 (Hf) 12.7 2.77 5.48 5 0.100 7.32 10 0.031 5.65 5 7.14 10 0.337 4.38 3 5.97 6 (As) 7.53 5.59 5 3.80 2 0.067 7.75 12 0.046 5.97 6 7.84 12 0.519 9.11 12 (Dp) 7.35 4.48 3 3.55 1 0.058 7.16 10 0.034 5.47 5 7.35 10 0.351 3.65 1 6.09 7 (Le) 3.43 6.14 7 0.74 0.029 4.57 3 0.341 0.38 8.48 12 0.025 9.21 12 1.60 (Lo) 2.80 6.58 8 0.040 5.91 6 0.311 8.71 12 0.071 9.11 12 (Ot) 25.4 6.58 8 0.076 7.31 10 0.025 5.70 6 6.98 9 0.284 5.50 5 5.45 5 (Ch) 5.32 5.11 4 2.33 0.043 6.35 7 0.026 4.97 4 6.58 8 0.383 3.43 1 6.16 7 (An) 2.81 5.96 6 0.17 0.062 6.98 9 0.013 6.60 8 4.45 3 0.253 6.04 6 4.39 3 (Ob) 6.54 4.22 2 2.74 0.045 6.32 7 0.019 6.73 8 5.48 5 0.331 4.63 3 5.18 4 (Sc) 3.27 6.18 7 0.22 0.044 5.31 5 0.021 5.65 5 5.62 5 0.141 6.92 9 1.98 (Eg) 0.79 6.49 8 2.50 0.000 0.000 8.71 12 0.002 8.18 12 1.84 R rted 0.0 0.2 0.4 0.6 0.8 1.0 3: R rted Eg Twitter 5.4 Twitter R rt R reply R rted 3 R rted 1 R rted 3 Lo Le R rted 1 R tw R rted 3 R rted Lo Le R rted R reply = N reply / N tweet (4) R reply 4 R reply 1 R reply 1 R tw R reply 3 R reply As 2 1 Lo Processing Society of Japan 275
R reply 0.0 0.2 0.4 0.6 0.8 1.0 4: R reply 6. [16], [17] 7. Twitter 13 100 [1] Burger, J. D., Henderson, J., Kim, G. and Zarrella, G.: Discriminating Gender on Twitter, Proceedings of the Conference on Empirical Methods in Natural Language Processing, pp. 1301 1309 (2011). [2] Golbeck, J., Robles, C., Edmondson, M. and Turner, K.: Predicting Personality from Twitter, 2011 IEEE Third International Conference on Privacy, Security, Risk and Trust (PASSAT) and 2011 IEEE Third Inernational Conference on Social Computing (SocialCom), IEEE, pp. 149 156 (2011). [3] Golbeck, J., Robles, C. and Turner, K.: Predicting Personality with Social Media, Proceedings of the 2011 Annual Conference Extended Abstracts on Human Factors in Computing Systems - CHI EA 11, New York, New York, USA, ACM Press, pp. 253 262 (2011). [4] Hum, N. J., Chamberlin, P. E., Hambright, B. L., Portwood, A. C., Schat, A. C. and Bevan, J. L.: A picture is worth a thousand words: A content analysis of Facebook profile photographs, Computers in Human Behavior, Vol. 27, No. 5, pp. 1828 1833 (2011). [5] Ikeda, K., Hattori, G., Ono, C., Asoh, H. and Higashino, T.: Twitter user profiling based on text and community mining for market analysis, Knowledge-Based Systems, Vol. 51, pp. 35 47 (2013). [6] Kapidzic, S.: Narcissism as a predictor of motivations behind Facebook profile picture selection., Cyberpsychology, Behavior and Social Networking, Vol. 16, No. 1, pp. 14 19 (2013). [7] Kim, D., Jo, Y., Moon, I.-C. and Oh, A.: Analysis of Twitter Lists as a Potential Source for Discovering Latent Characteristics of Users, ACM CHI Workshop on Microblogging, p. 4 (2010). [8] Pennacchiotti, M. and Gurumurthy, S.: Investigating Topic Models for Social Media User Recommendation, Proceedings of the 20th International Conference Companion on World Wide Web, New York, New York, USA, ACM Press, pp. 101 102 (2011). [9] Pennacchiotti, M. and Popescu, A.-m.: A Machine Learning Approach to Twitter User Classification, Proceedings of the Fifth International AAAI Conference on Weblogs and Social Media, pp. 281 288 (2010). [10] Qiu, L., Lin, H., Ramsay, J. and Yang, F.: You are what you tweet: Personality expression and perception on Twitter, Journal of Research in Personality, Vol. 46, No. 6, pp. 710 718 (2012). [11] Quercia, D., Kosinski, M., Stillwell, D. and Crowcroft, J.: Our Twitter Profiles, Our Selves: Predicting Personality with Twitter, 2011 IEEE Third International Conference on Privacy, Security, Risk and Trust (PASSAT) Processing Society of Japan 276
and 2011 IEEE Third Inernational Conference on Social Computing (SocialCom), IEEE, pp. 180 185 (2011). [12] Ramage, D., Dumais, S. and Liebling, D.: Characterizing Microblogs with Topic Models, Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media, pp. 130 137 (2010). [13] Siibak, A.: Constructing masculinity on a social networking site, Young, Vol. 18, No. 4, pp. 403 425 (2010). [14] Siibak, A. and Hernwall, P.: Looking like my favourite Barbie Online Gender Construc- tion of Tween Girls in Estonia and in Sweden Studies of Transition States and Societies, Vol. 3, No. 2, pp. 57 68 (2011). [15] Suler, J.: The Psychology of Avatars and Graphical Space in Multimedia Chat Communities or : How I Learned to Stop Worrying and Love My Palace Props, Chat communication, pp. 305 344 (1999). [16] Zhao, C. and Jiang, G.: Cultural differences on visual self-presentation through social networking site profile images, Proceedings of the 2011 Annual Conference on Human Factors in Computing Systems - CHI 11, New York, New York, USA, ACM Press, pp. 1129 1132 (2011). [17] Ziller, R. C.: Photographing the self: Methods for observing personal orientations., Sage publications, Inc (1990). Processing Society of Japan 277