Predicting popularity via a generative model with adaptive peeking window

Improving News Popularity Estimation via Weak Supervision and Meta-active Learning

“…Alternatively, previous studies [16]- [18] have focused on feature engineering as one of the challenges that face popularity estimation. For example, authors in [16] applied vocabulary clustering to social content to detect similar patterns of popular topics.…”

“…Another research [17] presents a preliminary analysis of content popularity before developing a regression model that employs the analysis results to predict popular trends in the future. Moreover, Bao et al [18] proposed a method that observes mobile social content to decide on the most significant attributes to build the final feature-driven model. However, most of these approaches are content-specific.…”

“…However, most of these approaches are content-specific. For example, they focus on certain types of content, such as videos [16] and tweets [18]. Therefore, the final models are restricted to analyze content history within a single observed domain.…”

Information cascades prediction with attention neural network

“…Yu et al [21] proposed a novel NEtworked Weibull Regression model for modeling microbehavioral dynamics that significantly improved the interpretability and generalizability of traditional survival models. Bao et al [23] modeled the popularity dynamics of the tweet in Twitter using the Hawkes process. They also proposed a method for exploring an adaptive peeking window for each tweet, which can synthesize all of the global dynamic information within the observed period into the predicted peek point.…”

“…Reflected in the time dimension is the aggregation of adoption events, which is also called bursting diffusion of the message. In our previous work [23], we demonstrated that different parts of the diffusion history have diverse influences on the future cascade size, and we proposed a method for obtaining the most effective part of the history to make an accurate prediction. Analogically, the pooling weights for the temporal property of different adoption events are automatically learned based on the structural embedding of the cascade graph g t to optimize the prediction of cascade growth.…”