I. Introduction
Data-driven models have become the default choice for building personalized recommendation services [1], [2]. These models typically focus on the correlation between item attributes and user feedback, suffering from the confounding bias [3], [4]. The source of such bias is the confounder that affects item attributes and user feedback simultaneously, leading to spurious correlations [5], [6]. For instance, the high quality of an item is the driving factor behind its high price, and it also tends to generate more positive ratings from users, resulting in a spurious correlation between high price and high rating. Fitting the data solely based on this correlation can lead to the over-recommendation of high-price items. Worse still, the confounding effect will hurt the fairness across item producers and make the model vulnerable to be attacked, e.g., some producers may intentionally increase the price for more exposure opportunities. It is thus essential to mitigate the confounding effect in recommendation.