Research on Discovery of Business Community Based on Ant Colony Algorithm

摘要

The network can offer the opportunity of further level and potential development for the e-commerce with different topics, different interests and different functions. Therefore, the research of network community discovery in E-commerce plays the most important role in the E-commerce. The ABDA (Ant-based Data Analysis) is based on the behavior of ant foraging and structuring graves, on the basis of ant colony clustering and optimization. First, we project the data source to be processed in the database onto a two-dimensional plane. Then, it will produce some virtual ants on the plane clustering analysis to the data source. Each data has a pheromone logo, by calculating the similarity of pheromone to realize clustering. According to the pheromone concentration, ants choose the optimal path to clustering. After finite iterations, the clustering results are obtained. The ABDA algorithm can not only realize high efficient, parallel, fast clustering, but also cluster thoroughly, considering the noise problem and doing a good deal with the outliers.