Predicting user attitudes toward smartphone ads using support vector machine

KangWoo Lee, Hyunseung Choo

International Journal of Mobile Communications, August 2015 (IF:1.221)


This study presents a computational model of smartphone ads that uses support vector machine (SVM). The model is used to simulate the well-known social phenomenon of ‘similarity attraction,’ which we analysed using both regression and pattern classification models. Smartphone call patterns were used to predict user personality for the given smartphone call patterns and ad types (extrovert or introvert), the model simulated the similarity attraction effect and predicted user attitudes toward the smartphone ad in terms of likeability, credibility and buying intention. The results indicated that the SVM model is a powerful tool for simulating similarity attraction and correctly classifies user attitude. The computational implication of the model is discussed in terms of customisation and persuasiveness.



Computational advertising, Smartphone call patterns, Support vector machine, prediction, Similarity attraction


View Full Text