Study On Recommendation For Specific-Oriented E-Commerce Services Via Open API

With the prosperity of e-commerce, more and more people are involved into e-commerce activities. Complex transactions and social relationships are formed among users. The ideas of Web2.0 such as openness, collaboration have been widely accepted. E-commerce platforms have proposed their own open strategies and open API to independent third party developers. On one hand third parties can develop value added service applications and develop program by introducing open API into independent websites, on the other hand it provides data sources for researchers.Definition of API is described. eBay developers program and Taobao Open Platform strategries are introduced. Deep comparison of TOP API and eBay API from five aspects are made. Using eBay open API, two different data extract solutions are given from the perspectives of items and users respectively to extract real, dynamic transaction from eBay website. Definition of e-commerce transaction network is made and weighted transaction network split algorithm is designed. Network analysis and visualization program Pajek are used to analyze splited networks. The complex adaptive features of E-commerce transaction network are analyzed in the fourth part. The simulation system of e-commerce transaction network is established using the extracted eBay transaction data based on Repast modelling and simulation platform. Different credit evaluation methods are compared using the transaction network. E-commerce service recommendation prototype system is designed and introduced.The transaction data is extracted from e-commerce platforms using open API. Multi-agent modelling and simulation method and complex network analysis technology are used to research e-commerce transaction network. The simulated system is used to compare different credit evaluation methods to discover potential knowledge and provide personalized information recommendation services.