| Summary: | 碩士 === 輔仁大學 === 資訊管理學系 === 95 === Electronic commerce (e-commerce) has grown dramatically over the years into a business model of vital importance in the 21st century. Yet, consumers still encounter problems such as false expectation and purchase disputes while shopping online. Moreover, the buyer decision-making process is often prolonged due to the lack of purchasing assistance on the electronic commerce platforms’ end therefore wasting consumers’ time and energy. Furthermore, the inability of majority of online stores to understand consumers’ needs creates obstacles in providing personalized products and service efficiently.
This research’s goal is to design an intelligent system based on Web2.0’s fundamental structure to assign objective, pluralistic and multilevel product information to goods through collective intelligence. The innovative Tag Vectorial Model and Folksonomy-based Ranking Algorithm are proposed to combine with Collaboration Recommendation Rules to capture user preference in providing more accurate personalized products and service.
Results show, the increase of the numbers of searches and ratings conducted improves the system’s accuracy in capturing user preference and in turn allows the system to present more satisfactory product search results to consumers, effectively assisting them to narrow down product choices thus shortening the buyer decision-making process. It is also shown, the system frequently recommends products surprising to consumers. At last, after analyzing Tag Vectorial Model with graphs, it is discovered that Tag Vectorial Model is indeed able to present consumers’ motivations. This may help online businesses to further understand their consumers and provide personalized products and service to them. Subsequently, new business opportunities may be created.
|
|---|
