| Summary: | 碩士 === 國立中央大學 === 企業管理學系 === 105 === With the growing popularity of the social network, the number of people using the social network to communicate and interactive with others increased steadily. As a result, social commerce has become a new phenomena.
In the past, most of the product recommendation in Microblog only deal with personal preferences and interests, and ignores other possible factors such as crowd Interest, Popularity of products, reputation of creators, types of preference and recency. These variables are used by facebook to recommend posts to users. Therefore, this research adapted the five aspects and analyzed their effectiveness to recommend products on social media sites.
The empirical results show that the Interest, Popularity and Type have significant impacts on recommendation effectivness. In addition, this studies also utilized Artificial Neural Networks to predict the click through rates of recommended web pages. The results show that the Artificial Neural Networks have better predictive effect then Linear Regression. However, the three variables identified by Linear Regression indeed outperform the other variables.
|
|---|
