【摘要】 为揭示产品评论数量与购买数量间的关系,采用数据驱动的符号回归方法,该方法不需预先假设模型结构,从数据中自动学习出描述产品评论数量与购买数量的关系模型和相关参数.数据采自天猫商城网站中隶属于30种商品的5 387个售出产品以及相关评论,利用符号回归方法进行自动学习,发现了在实证方法中假设的线性模型.这种模式在购买数量与评论数量之间的关系上具有更高的适应性.更多还原

【Abstract】 In order to reveal the relationship between the volume of purchases and the volume of reviews,this paper introduces a data-driven symbolic regression method to discover both the form of the model and its parameters intelligently based on a large amount of data. Compared with previous empirical studies, symbolic regression via genetic programming does not need a prior hypothetical model form to fit the observed data.The data sets of this paper are collected from Tmall.com. The data sets include 5 387 products belonging to 30 categories. By the method, the linear model assumed in the prior empirical methods is found. Meanwhile, new nonlinear models are also found. Such models hasve a better fitness in terms of the relationship between the volume of purchases and the volume of reviews.更多还原