Almost all existing causal feature selection methods are proposed without considering the problem of sample selection bias. However, in practice, as data-gathering process cannot be fully controlled, sample selection bias often occurs, leading to spurious correlations between features and the class variable, which seriously deteriorates the performance of those existing methods. In this article, we study the problem of causal feature selection under sample selection bias and propose a novel Progressive Causal Feature Selection (PCFS) algorithm which has three phases. First, PCFS learns the sample weights to balance the treated group and control group distributions corresponding to each feature for removing spurious correlations. Second, based on the sample weights, PCFS uses a weighted cross-entropy model to estimate the causal effect of each feature and removes some irrelevant features from the confounder set. Third, PCFS progressively repeats the first two phases to remove more irrelevant features and finally obtains a causal feature set. Using synthetic and real-world datasets, the experiments have validated the effectiveness of PCFS, in comparison with several state-of-the-art classical and causal feature selection methods.
Labels provide a quick and effective solution to obtain people interesting content from large‐scale social network information. The current interest label extraction method based on the subgraph stream proves the feasibility of the subgraph stream for user label extraction. However, it is extremely time‐consuming for constructing subgraphs. As an effective mathematical method to deal with fuzzy and uncertain information, rough set‐based representations for subgraph stream construction are capable of capturing the uncertainties of the social network. Therefore, we propose an effective approach called RS_UNITE_SS (namely, rough‐set‐based user‐networked interest topic extraction in the form of subgraph stream), which is suitable for large‐scale social network user interest label extraction. Specifically, we first propose the subgraph division algorithm to construct a subgraph stream by incorporating a rough set. Then, the algorithm for user real‐time interest label extraction based on upper approximation (RILE) is proposed by using sequentially characteristics of the subgraph. Empirically, we evaluate RS_UNITE_SS over real‐world datasets, and experimental results demonstrate that our proposed approach is more computationally efficient than existing methods while achieving higher precision value and MRR value.