Abstract: 【Purpose/significance】Mining the cognitive process of e-commerce users' information organization and understanding, provid‑ ing the basis for enterprises to deeply understand users' information processing and decision-making process, optimize the way of com‑ modity information presentation and service, and enhance user experience and marketing effect【. Method/process】Based on the hierarchi‑ cal factor graph to construct the information cognitive structure model of e-commerce users, based on the behavioral data of e-commerce users' information episodic processing on the product page browsing, the variational reasoning algorithm is used to deduce the message transfer path and relationship between users' a priori knowledge and the processing of new knowledge, and to excavate the user's informa‑ tion cognitive structure【. Result/conclusion】The experimental results show that the cognitive structure mining method based on hierarchi‑ cal factor graph can effectively identify the information cognitive structure of e-commerce users from China, the Philippines and Malay‑ sia. By comparing the information cognitive structure of users in different countries, it is found that there are certain differences in infor‑ mation processing and decision making among users in different cultural backgrounds.【Innovation/limitation】The study reveals the con‑ struction process of user information cognitive structure, and in the future, we will deeply explore the integration of the user cognitive structure model with the context-awareness model, and introduce contextual information as an input parameter into the analysis of user cognitive structure to more accurately understand and predict the cognitive structure and behavioral patterns of users in different contexts.