Generative Models of Information Diffusion with Asynchronous Timedelay

Kazumi Saito, Masahiro Kimura, Kouzou Ohara, Hiroshi Motoda
Proceedings of 2nd Asian Conference on Machine Learning, PMLR 13:193-208, 2010.

Abstract

We address the problem of formalizing an information diffusion process on a social network as a generative model in the machine learning framework so that we can learn model parameters from the observation. Time delay plays an important role in formulating the likelihood function as well as for the analyses of information diffusion. We identified that there are two different types of time delay: link delay and node delay. The former corresponds to the delay associated with information propagation, and the latter corresponds to the delay due to human action. We further identified that there are two distinctions of the way the activation from the multiple parents is updated: nonoverride and override. The former sticks to the initial activation and the latter can decide to update the time to activate multiple times. We formulated the likelihood function of the well known diffusion models: independent cascade and linear threshold, both enhanced with asynchronous time delay distinguishing the difference in two types of delay and two types of update scheme. Simulation using four real world networks reveals that there are differences in the spread of information diffusion and they strongly depend on the choice of the parameter values and the denseness of the network.

Cite this Paper

BibTeX
@InProceedings{pmlr-v13-saito10a, title = {Generative Models of Information Diffusion with Asynchronous Timedelay}, author = {Saito, Kazumi and Kimura, Masahiro and Ohara, Kouzou and Motoda, Hiroshi}, booktitle = {Proceedings of 2nd Asian Conference on Machine Learning}, pages = {193--208}, year = {2010}, editor = {Sugiyama, Masashi and Yang, Qiang}, volume = {13}, series = {Proceedings of Machine Learning Research}, address = {Tokyo, Japan}, month = {08--10 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v13/saito10a/saito10a.pdf}, url = {https://proceedings.mlr.press/v13/saito10a.html}, abstract = {We address the problem of formalizing an information diffusion process on a social network as a generative model in the machine learning framework so that we can learn model parameters from the observation. Time delay plays an important role in formulating the likelihood function as well as for the analyses of information diffusion. We identified that there are two different types of time delay: link delay and node delay. The former corresponds to the delay associated with information propagation, and the latter corresponds to the delay due to human action. We further identified that there are two distinctions of the way the activation from the multiple parents is updated: nonoverride and override. The former sticks to the initial activation and the latter can decide to update the time to activate multiple times. We formulated the likelihood function of the well known diffusion models: independent cascade and linear threshold, both enhanced with asynchronous time delay distinguishing the difference in two types of delay and two types of update scheme. Simulation using four real world networks reveals that there are differences in the spread of information diffusion and they strongly depend on the choice of the parameter values and the denseness of the network.} }
Endnote
%0 Conference Paper %T Generative Models of Information Diffusion with Asynchronous Timedelay %A Kazumi Saito %A Masahiro Kimura %A Kouzou Ohara %A Hiroshi Motoda %B Proceedings of 2nd Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2010 %E Masashi Sugiyama %E Qiang Yang %F pmlr-v13-saito10a %I PMLR %P 193--208 %U https://proceedings.mlr.press/v13/saito10a.html %V 13 %X We address the problem of formalizing an information diffusion process on a social network as a generative model in the machine learning framework so that we can learn model parameters from the observation. Time delay plays an important role in formulating the likelihood function as well as for the analyses of information diffusion. We identified that there are two different types of time delay: link delay and node delay. The former corresponds to the delay associated with information propagation, and the latter corresponds to the delay due to human action. We further identified that there are two distinctions of the way the activation from the multiple parents is updated: nonoverride and override. The former sticks to the initial activation and the latter can decide to update the time to activate multiple times. We formulated the likelihood function of the well known diffusion models: independent cascade and linear threshold, both enhanced with asynchronous time delay distinguishing the difference in two types of delay and two types of update scheme. Simulation using four real world networks reveals that there are differences in the spread of information diffusion and they strongly depend on the choice of the parameter values and the denseness of the network.
RIS
TY - CPAPER TI - Generative Models of Information Diffusion with Asynchronous Timedelay AU - Kazumi Saito AU - Masahiro Kimura AU - Kouzou Ohara AU - Hiroshi Motoda BT - Proceedings of 2nd Asian Conference on Machine Learning DA - 2010/10/31 ED - Masashi Sugiyama ED - Qiang Yang ID - pmlr-v13-saito10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 13 SP - 193 EP - 208 L1 - http://proceedings.mlr.press/v13/saito10a/saito10a.pdf UR - https://proceedings.mlr.press/v13/saito10a.html AB - We address the problem of formalizing an information diffusion process on a social network as a generative model in the machine learning framework so that we can learn model parameters from the observation. Time delay plays an important role in formulating the likelihood function as well as for the analyses of information diffusion. We identified that there are two different types of time delay: link delay and node delay. The former corresponds to the delay associated with information propagation, and the latter corresponds to the delay due to human action. We further identified that there are two distinctions of the way the activation from the multiple parents is updated: nonoverride and override. The former sticks to the initial activation and the latter can decide to update the time to activate multiple times. We formulated the likelihood function of the well known diffusion models: independent cascade and linear threshold, both enhanced with asynchronous time delay distinguishing the difference in two types of delay and two types of update scheme. Simulation using four real world networks reveals that there are differences in the spread of information diffusion and they strongly depend on the choice of the parameter values and the denseness of the network. ER -
APA
Saito, K., Kimura, M., Ohara, K. & Motoda, H.. (2010). Generative Models of Information Diffusion with Asynchronous Timedelay. Proceedings of 2nd Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 13:193-208 Available from https://proceedings.mlr.press/v13/saito10a.html.

Related Material