基于图排序的社会媒体用户的消费意图检测
摘要
Consumption intention in a microblog indicates either an imminent or a future purchase, and this indicates a buying intention. Consumption-intention detection aims to determine the willingness of a user to make a purchase. The consumption intention in a microblog should contain two important elements, namely the trigger words and the consumption intention target (the product to be bought). These two elements directly indicate a user's purchase intention, which is an important feature of the user's consumption intention. In this paper, we propose to study the problem of detecting consumption intention in microblogs based on graph ranking. This method can be applied to cases where there is a large amount of data, the amount of labeled data is relatively small, and all data can be involved in the learning process of the graph-ranking algorithm. Experimental results show that the graph-ranking-based method is effective for detecting consumption intention.