One size does not fit all : profiling personalized time-evolving user behaviors
- Author
- Pravallika Devineni, Evangelos E. Papalexakis, Danai Koutra, A. Seza Doğruöz (UGent) and Michalis Faloutsos
- Organization
- Abstract
- Given the set of social interactions of a user, how can we detect changes in interaction patterns over time? While most previous work has focused on studying network-wide properties and spotting outlier users, the dynamics of individual user interactions remain largely unexplored. This work sets out to explore those dynamics in a way that is minimally invasive to privacy, thus, avoids to rely on the textual content of user posts---except for validation. Our contributions are two-fold. First, in contrast to previous studies, we challenge the use of a fixed interval of observation. We introduce and empirically validate the "Temporal Asymmetry Hypothesis", which states that appropriate observation intervals should vary both among users and over time for the same user. We validate this hypothesis using eight different datasets, including email, messaging, and social networks data. Second, we propose iNET, a comprehensive analytic and visualization framework which provides personalized insights into user behavior and operates in a streaming fashion. iNET learns personalized baseline behaviors of users and uses them to identify events that signify changes in user behavior. We evaluate the effectiveness of iNET by analyzing more than half a million interactions from Facebook users. Labeling of the identified changes in user behavior showed that iNET is able to capture a wide spectrum of exogenous and endogenous events, while the baselines are less diverse in nature and capture only 66% of that spectrum. Furthermore, iNET exhibited the highest precision (95%) compared to all competing approaches.
- Keywords
- social networks, interaction patterns, social media, Lt3, computational social sciences
Downloads
-
Dogruoz IEEEACM17 socialnetworks.pdf
- full text (Accepted manuscript)
- |
- open access
- |
- |
- 1.26 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8694737
- MLA
- Devineni, Pravallika, et al. “One Size Does Not Fit All : Profiling Personalized Time-Evolving User Behaviors.” Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017, edited by Jana Diesner et al., Association for Computing Machinery (ACM), 2017, pp. 331–40, doi:10.1145/3110025.3110050.
- APA
- Devineni, P., Papalexakis, E. E., Koutra, D., Doğruöz, A. S., & Faloutsos, M. (2017). One size does not fit all : profiling personalized time-evolving user behaviors. In J. Diesner, E. Ferrari, & G. Xu (Eds.), Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017 (pp. 331–340). https://doi.org/10.1145/3110025.3110050
- Chicago author-date
- Devineni, Pravallika, Evangelos E. Papalexakis, Danai Koutra, A. Seza Doğruöz, and Michalis Faloutsos. 2017. “One Size Does Not Fit All : Profiling Personalized Time-Evolving User Behaviors.” In Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017, edited by Jana Diesner, Elena Ferrari, and Guandong Xu, 331–40. New York: Association for Computing Machinery (ACM). https://doi.org/10.1145/3110025.3110050.
- Chicago author-date (all authors)
- Devineni, Pravallika, Evangelos E. Papalexakis, Danai Koutra, A. Seza Doğruöz, and Michalis Faloutsos. 2017. “One Size Does Not Fit All : Profiling Personalized Time-Evolving User Behaviors.” In Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017, ed by. Jana Diesner, Elena Ferrari, and Guandong Xu, 331–340. New York: Association for Computing Machinery (ACM). doi:10.1145/3110025.3110050.
- Vancouver
- 1.Devineni P, Papalexakis EE, Koutra D, Doğruöz AS, Faloutsos M. One size does not fit all : profiling personalized time-evolving user behaviors. In: Diesner J, Ferrari E, Xu G, editors. Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017. New York: Association for Computing Machinery (ACM); 2017. p. 331–40.
- IEEE
- [1]P. Devineni, E. E. Papalexakis, D. Koutra, A. S. Doğruöz, and M. Faloutsos, “One size does not fit all : profiling personalized time-evolving user behaviors,” in Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017, Sydney, Australia, 2017, pp. 331–340.
@inproceedings{8694737,
abstract = {{Given the set of social interactions of a user, how can we detect changes in interaction patterns over time? While most previous work has focused on studying network-wide properties and spotting outlier users, the dynamics of individual user interactions remain largely unexplored. This work sets out to explore those dynamics in a way that is minimally invasive to privacy, thus, avoids to rely on the textual content of user posts---except for validation. Our contributions are two-fold. First, in contrast to previous studies, we challenge the use of a fixed interval of observation. We introduce and empirically validate the "Temporal Asymmetry Hypothesis", which states that appropriate observation intervals should vary both among users and over time for the same user. We validate this hypothesis using eight different datasets, including email, messaging, and social networks data. Second, we propose iNET, a comprehensive analytic and visualization framework which provides personalized insights into user behavior and operates in a streaming fashion. iNET learns personalized baseline behaviors of users and uses them to identify events that signify changes in user behavior. We evaluate the effectiveness of iNET by analyzing more than half a million interactions from Facebook users. Labeling of the identified changes in user behavior showed that iNET is able to capture a wide spectrum of exogenous and endogenous events, while the baselines are less diverse in nature and capture only 66% of that spectrum. Furthermore, iNET exhibited the highest precision (95%) compared to all competing approaches.}},
author = {{Devineni, Pravallika and Papalexakis, Evangelos E. and Koutra, Danai and Doğruöz, A. Seza and Faloutsos, Michalis}},
booktitle = {{Proceedings of the 2017 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2017}},
editor = {{Diesner, Jana and Ferrari, Elena and Xu, Guandong}},
isbn = {{9781450349932}},
keywords = {{social networks,interaction patterns,social media,Lt3,computational social sciences}},
language = {{eng}},
location = {{Sydney, Australia}},
pages = {{331--340}},
publisher = {{Association for Computing Machinery (ACM)}},
title = {{One size does not fit all : profiling personalized time-evolving user behaviors}},
url = {{http://doi.org/10.1145/3110025.3110050}},
year = {{2017}},
}
- Altmetric
- View in Altmetric