Anzeige
Erschienen in:
2017 | OriginalPaper | Buchkapitel
Temporal Prediction Model for Social Information Propagation
verfasst von : Fei Teng, Rong Tang, Yan Yang, Hongjie Wang, Rongjie Dai
Erschienen in: Rough Sets
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Prediction of information propagation is an important issue in research of social network. Recent researches can be divided into graph or non-graph approaches. Most of non-graph approaches use regression analysis and probability model, seldomly considering clustering features of social time series. In clustering-based temporal prediction model, every cluster center is treated as a propagation pattern, and so that the prediction can be realized through classification to find out the nearest-neighbor pattern. Prediction performance may be influenced by clustering performance based on clustering approaches. This paper proposes a new model Scaling Clustering based Temporal Prediction Model (SCTPM), which is applicable for predicting propagation pattern of social information. Through 10-fold cross-validation experiments on twitter and phrase datasets, SCTPM obtains lower prediction bias and variance than the existing clustering-based models.