建議先修過程式設計、資料探勘/文字探勘、機器學習與陳建錦合授

Since the advent of Web 2.0 and online social networking, social media platforms, establishing and fostering general-purpose or interest-based communities that give their users the power to actively create and contribute user-generated contents and connect and interact with others, have exploded in popularity. As a result, social media has become one unique, novel source of big data, providing great opportunities and magnificent potential for research to analyze and understand human behavioral patterns or to develop advanced analytics techniques that analyze social media data for business decision support or more effective social media management. For example, individuals increasingly share on social media platforms their experiences with, preferences for, and opinions about a wide range of products, entities (e.g., organizations, celebrities, politicians), emerging events, public policies, and so on. Organizations increasingly rely on these user reviews to answer such questions as: What do customers say about us and about our competitors? What do they like or dislike about our products? What causes customers to become dissatisfied?

Social media analytics (SMA) refers to “the process and methods of collecting and analyzing data gathered from social media channels to support business decisions.” The objective of this course is to help students understand commonly discussed topics and their corresponding analytics methods related to social media analytics. This course is structured into three modules: SMA essentials, network-based SMA methods, and text-based SMA methods. Technically speaking, social media analytics is a confluence of research in data mining, text mining, and social network analysis. Thus, in the “SMA essentials” module, we will cover these fundamental building blocks of social media analytics, including 1) data mining essentials, 2) text mining essentials, and 3) social network analysis and link prediction. The “network-based SMA methods” module will cover the topics that heavily rely on the structural analysis of social networks to support the analysis of social media. They include 1) community detection, 2) influence modeling, and 3) opinion leader detection. In the “text- based SMA methods” module, we will discuss the methods and applications that predominantly exploit the texts (user-generated contents) shared on social media platforms. They include 1) sentiment analysis, 2) social media for marketing intelligence, 3) user profiling, and 4) fake review detection.