Enterprise marketing strategy using big data mining technology combined with XGBoost model in the new economic era

• Published in: PloS one 2023-06, Vol.18 (6), p.e0285506-e0285506
• Main Author: Chen, Huijun
• Format: Article
• Language: English
• Subjects: Algorithms; Artificial intelligence; Big Data; Biology and Life Sciences; Computer and Information Sciences; Consumption; Customer relationship management; Customers; Data analysis; Data Mining; Decision making; Design; Evaluation; Innovations; Internet; Knowledge discovery; Literature reviews; Machine learning; Market strategy; Marketing; Medical device industry; Medical equipment; Medical research; Network computers; Physical Sciences; Research and Analysis Methods; Social Sciences; Society; Technology; Technology application
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0285506

The technological development in the new economic era has brought challenges to enterprises. Enterprises need to use massive and effective consumption information to provide customers with high-quality customized services. Big data technology has strong mining ability. The relevant theories of computer data mining technology are summarized to optimize the marketing strategy of enterprises. The application of data mining in precision marketing services is analyzed. Extreme Gradient Boosting (XGBoost) has shown strong advantages in machine learning algorithms. In order to help enterprises to analyze customer data quickly and accurately, the characteristics of XGBoost feedback are used to reverse the main factors that can affect customer activation cards, and effective analysis is carried out for these factors. The data obtained from the analysis points out the direction of effective marketing for potential customers to be activated. Finally, the performance of XGBoost is compared with the other three methods. The characteristics that affect the top 7 prediction results are tested for differences. The results show that: (1) the accuracy and recall rate of the proposed model are higher than other algorithms, and the performance is the best. (2) The significance p values of the features included in the test are all less than 0.001. The data shows that there is a very significant difference between the proposed features and the results of activation or not. The contributions of this paper are mainly reflected in two aspects. 1. Four precision marketing strategies based on big data mining are designed to provide scientific support for enterprise decision-making. 2. The improvement of the connection rate and stickiness between enterprises and customers has played a huge driving role in overall customer marketing.