Latent Factors Meet Homophily in Diffusion Modelling
Diffusion is an important dynamics that helps spreading information within an online social network. While there are already numerous models for single item diffusion, few have studied diffusion of multiple items, especially when items can interact with o
- PDF / 490,639 Bytes
- 18 Pages / 439.37 x 666.142 pts Page_size
- 20 Downloads / 175 Views
Abstract. Diffusion is an important dynamics that helps spreading information within an online social network. While there are already numerous models for single item diffusion, few have studied diffusion of multiple items, especially when items can interact with one another due to their inter-similarity. Moreover, the well-known homophily effect is rarely considered explicitly in the existing diffusion models. This work therefore fills this gap by proposing a novel model called Topic level Interaction Homophily Aware Diffusion (TIHAD) to include both latent factor level interaction among items and homophily factor in diffusion. The model determines item interaction based on latent factors and edge strengths based on homophily factor in the computation of social influence. An algorithm for training TIHAD model is also proposed. Our experiments on synthetic and real datasets show that: (a) homophily increases diffusion significantly, and (b) item interaction at topic level boosts diffusion among similar items. A case study on hashtag diffusion in Twitter also shows that TIHAD outperforms the baseline model in the hashtag adoption prediction task.
1
Introduction
Ubiquitous presence of online social networks (OSN) has made information diffusion an important topic that attracts much research interests. While many items may diffuse in a social network simultaneously, most existing models of diffusion are built upon independent contagion assumption whereby the diffusion of each item is assumed (at least implicitly) to happen independent of other items. The interaction among items during diffusion is thus left out of the picture. This is obviously not true in the complex dynamics of diffusion process. For instance, the diffusion of iPhones in the Facebook friendship network may interact favorably with that of iPad; and the diffusion of a catchy phrase on Twitter also aids the diffusion of its variants. Interaction Among Items. Modeling these interactions is crucial in both theory and practice since it helps us understand the detailed dynamics of multiple item diffusion. It is also valuable for business to develop suitable strategies to promote diffusion of their own items considering the other items that have been diffused recently or are being diffused. It may be good to time the diffusion of c Springer International Publishing Switzerland 2015 A. Appice et al. (Eds.): ECML PKDD 2015, Part II, LNAI 9285, pp. 701–718, 2015. DOI: 10.1007/978-3-319-23525-7 43
702
M.-D. Luu and E.-P. Lim
a new item with the diffusion of other similar items (possibly by the business or other businesses) to achieve a larger reach. This idea of diffusion with item interaction can be further illustrated in the following motivating example. Example. A user may be inspired to watch the movie version of “Hunger Games” after observing some neighbors already read the book. Moreover, if both the book and the movie versions were adopted by a neighbor, the user will even be more likely to adopt the movie than if only one of them was adopted by the neighbor (as he may be more
Data Loading...
