Digital Analysis of Information on Cross-Bordere-Commerceplatform Guide Platforms Based on Artificial Intelligence Algorithms

At a time when big data is increasingly developed and constantly updated, cross-bordere-commerceplatform also needs to keep up with the trend of data. Facing the shortcomings of traditional e-commerce, providing users with personalized shopping guide information and improving the credibility and operational efficiency of the platform are the new business models for cross-bordere-commerceplatform in the era of big data. However, due to the lack of accuracy of traditional algorithms in product recommendations, it is not possible to digitally process and provide personalized recommendations for a large range of guide platform information when faced with the increasing volume of users who are increasingly shopping across borders. So, it is important to research the data information of cross-bordere-commerceplatform and build predictive models that link the consumer’s behavioural data on the website with historical sales data in order to infer the consumer’s willingness to purchase the target product. In this thesis, we will parallelize two kinds of MapReduce and one kind of map on the platform to obtain users’ preferences. For the consumer’s hobby, association rule mining is used to obtain products that match the user’s hobby and make recommendations to the user. The study proves that the AI algorithm is more accurate than other algorithms and meets the requirements of big data processing. The CTR of commodity information as well as the conversion rate are the best results.