Predicting new adopters via socially-aware neural graph collaborative filtering
Computational Data and Social Networks: 8th International Conference, CSoNet …, 2019•Springer
We predict new adopters of specific items by proposing S-NGCF, a socially-aware neural
graph collaborative filtering model. This model uses information about social influence and
item adoptions; then it learns the representation of user-item relationships via a graph
convolutional network. Experiments show that social influence is essential for adopter
prediction. S-NGCF outperforms the prediction of new adopters compared to state-of-the-art
methods by 18%.
graph collaborative filtering model. This model uses information about social influence and
item adoptions; then it learns the representation of user-item relationships via a graph
convolutional network. Experiments show that social influence is essential for adopter
prediction. S-NGCF outperforms the prediction of new adopters compared to state-of-the-art
methods by 18%.
Abstract
We predict new adopters of specific items by proposing S-NGCF, a socially-aware neural graph collaborative filtering model. This model uses information about social influence and item adoptions; then it learns the representation of user-item relationships via a graph convolutional network. Experiments show that social influence is essential for adopter prediction. S-NGCF outperforms the prediction of new adopters compared to state-of-the-art methods by 18%.
Springer
Showing the best result for this search. See all results