| Summary: | In recent years, mobility empowered by smart phones, tablets and numerous applications running on those mobile devices is transforming the way people live and work in the digital age. Innovations and new business models are emerging that take advantages of this rise of mobile computing. Despite tremendous opportunities promised by the transition to mobility, challenges exist before its full potential can be realized by society as well as by companies. For example, the spread of real-time targeting technologies in mobile display advertising creates a new challenge of how to efficiently allocate countless categories of advertising opportunities, or impressions, in real time. For another example, social broadcasting networks such as Twitter in the U.S. and "Weibo" in China are making it extremely convenient for consumers to spread word-of-mouth (WOM) among them, which both poses new challenges and offers new opportunities to companies wishing to harness the power of consumer WOM. The dissertation contains three essays exploring those issues. In the first essay, the concept of "smart market" for impression allocation is proposed, which emphasizes allocation contingent on uncertain supply and promotes coordination among advertisers across impression categories. A new theory is developed to solve the complicated optimization problem, which leads to a "decomposition and standardization" algorithm. In the second essay, I investigated whether and how Twitter WOM affects movie sales by estimating a dynamic panel data model using publicly available data and well-known machine learning algorithms. I found that chatter on Twitter does matter; however, the magnitude and direction of the effect depends on whom the WOM is from and what the WOM is about. The findings provide new perspectives to understand the effect of WOM on product sales and have important managerial implications. The third essay examines the possibility of designing social-broadcasting-based business intelligence (BI) systems that utilizes real-time information extracted from social broadcasting networks with text mining techniques. A new framework is proposed for this purpose and a Twitter-based BI system is designed and implemented that forecasts movie box office revenues during the opening weekend and daily revenue four weeks after the release of a movie. Preliminary results suggest that social-broadcasting-based BI systems have great potential and are worth exploring by both researchers and practitioners. === text
|
|---|
