English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

One important task in the study of information cascade is to predict thefuture recipients of a message given its past spreading trajectory. While thenetwork structure serves as the backbone of the spreading, an accurateprediction can hardly be made without the knowledge of the dynamics on thenetwork. The temporal information in the spreading sequence captures manyhidden features, but predictions based on sequence alone have theirlimitations. Recent efforts start to explore the possibility of combining boththe network structure and the temporal feature. Here, we propose a newend-to-end prediction method CasSeqGCN in which the structure and temporalfeature are simultaneously taken into account. A cascade is divided intomultiple snapshots which record the network topology and the state of nodes.The graph convolutional network (GCN) is used to learn the representation of asnapshot. A novel aggregation method based on dynamic routing is proposed toaggregate node representation and the long short-term memory (LSTM) model isused to extract temporal information. CasSeqGCN predicts the future cascadesize more accurately compared with other state-of-art baseline methods. Theablation study demonstrates that the improvement mainly comes from the designof the input and the GCN layer. We explicitly design an experiment to show thequality of the cascade representation learned by our approach is better thanother methods. Our work proposes a new approach to combine the structural andtemporal features, which not only gives a useful baseline model for futurestudies of cascade prediction, but also brings new insights on a widecollection of problems related with dynamics on and of the network.

CasSeqGCN: Combining Network Structure and Temporal Sequence to Predict Information Cascades