The Study On Purchasers' Preference Of The Third-Party B2B Platform Based On Improved Ant Colony Clustering Algorithm

B2B e-commerce plays an important role in the Internet economy, which has an important influence on the current business activities. At present, an increasing number of small and medium enterprises choose the third-party B2B e-commerce platform for business activities. Confronted with the gradually increasing demand of the SMEs as well as the more and more intense competition between other platforms, the third-party B2B platform needs to further improve itself, promote service quality, and attract and retain more customers, in order to survive and develop in the market.According to the influencing factors of customers'decision process, analyzing the characteristics of different customers'preference and implementing differentiated management strategy is an effective way to improve the competitiveness of the third-party B2B platform. From the purchasers' the perspective, this thesis focuses on the enterprise purchasers and analyses their preference. The main contents of this thesis are as follows:Firstly, this thesis analyses the existing related research in depth, including the development of the third-party B2B platform, customers' preferences and other related concepts. Meanwhile, this thesis summarizes the relevant methods and technologies to obtain customers' preferences, analyzing the advantages and disadvantages of different algorithms, which lays the foundation for the following research.Secondly, this thesis selects the ant colony clustering algorithm as the main method of analysis of purchasers' preferences. After improving the algorithm, and the validity of the improved algorithm is verified. Based on the improved algorithm, the third-party B2B platform purchasers' preference analysis model is constructed. According to the three aspects of the purchasers, the third-party B2B platform and the suppliers, an index system to measure the preference of purchasers is also established on the basis of the model. This thesis clusters the purchasers through the ant colony clustering algorithm of the purchasers' preference analysis model, and analyses different purchaser's preference according to the index system. Basing on theoretical analysis, this thesis collects data and has carried on an empirical study, and proposes some management advice according to the results, which helps the platform to implement different management methods, achieving the goal of improving the competitive advantage of platform.Finally, through the above theoretical and empirical research, this thesis summarizes deficiencies of the research, puts forward the direction of development, and proposes some countermeasures and suggestions.