Summary: | 碩士 === 國立高雄應用科技大學 === 電機工程系博碩士班 === 102 === Social media platforms can be used to quickly and widely propagate important message. Due to the explosive growth of social-media users, the opinions regarding utilization experiences and quality reviews of some newly released products provided by consumers, and recommendation made by social-opinion leaders, can effectively influence public consumer’s purchasing behaviors and stimulate marketing demands by broadcasting social messages. Motivated by this, in this work we utilized the collected social messages as a corpus for experiments, and employed a neural network approach to perform text mining on the corpus containing social messages associated with reviews of specific products, and developed a quantitative evaluation technique to measure influences of social-opinion leaders. According to the experimental results on clustering product reviews, and combining the computation of social leaders’ influence indicators, the novel approach and analytics method developed in our work can identify the functionality of social media in implementing a supportive way for product marketing.
|