基于旅客社交网络中的家庭出行预测

Predicting the Family Travels Based on Passenger Social Networks

家庭是客运市场最常见的消费单元之一。如果能够对家庭单元在未来一段时间是否出行做出准确的预测,将有助于客运、旅游等相关行业为家庭旅客提供个性化的服务和相关产品推荐。预测一个家庭在未来一段时间是否会出行是一个分类问题。构建家庭的人口统计学特征、历史行为特征以及预测时间窗口特征作为分类属性,使用多种分类算法进行家庭出行的预测。还构建基于共同出行关系的旅客社交网络,通过构建家庭成员的社交网络属性来进一步描述家庭特征。在一个旅客历史数据集上进行实验,取得了较好的效果。

Family passenger groups are one of the most popular travel consumption units in the passenger transportation market. If we can make an accurate forecast that if a family will travel over the next period of time, it will help the passenger transportation, tourism and other related industries to provide personalized products or services for families. Family travel prediction is defined as a classification problem. The solution is constructing family＇s demographic characteristics, historical behavior characteristics and predicting time window features for a classifier to forecast family travels. We also construct passenger social networks by extracting co-travel relations between passengers from their historical travel records and generate the social network features of family members to further describe the family characteristics. Experiments on a historical passenger travel data set show the efficiency of our method.