Characterizing user interest in NoSQL databases of social question and answer data

With the advent of social media technology for sharing commonly asked questions and answers among end users, there is rapidly growing interest in understanding the characteristics as well as utilizing social question and answer (QA) data. Not only SQL (NoSQL) is a popular technical topic on social question answering Web sites and is gaining popularity with emerging demands for scalable databases of big data. Despite the great interest of users in NoSQL technology, an attempt to analyze how the actual users react to NoSQL has not yet been made. Thus, in the present work, we utilize the QA data acquired from Stack Overflow (a QA Web site that works as a large knowledge repository) to understand how people perceive NoSQL technology. To this end, latent Dirichlet allocation (LDA) topic modeling techniques are used to discover the trend of NoSQL databases. In addition, we examine a weighted LDA model to reflect the influence of answers and finally propose the topic discrimination value in an attempt to find topics that distinguish each NoSQL database.