Doctor of Philosophy (PhD)
Department of Computer Science and Engineering
Mobile application (app) stores, such as Google Play and the Apple App Store, have recently emerged as a new model of online distribution platform. These stores have expanded in size in the past five years to host millions of apps, offering end-users of mobile software virtually unlimited options to choose from. In such a competitive market, no app is too big to fail. In fact, recent evidence has shown that most apps lose their users within the first 90 days after initial release. Therefore, app developers have to remain up-to-date with their end-users’ needs in order to survive. Staying close to the user not only minimizes the risk of failure, but also serves as a key factor in achieving market competitiveness as well as managing and sustaining innovation. However, establishing effective communication channels with app users can be a very challenging and demanding process. Specifically, users' needs are often tacit, embedded in the complex interplay between the user, system, and market components of the mobile app ecosystem. Furthermore, such needs are scattered over multiple channels of feedback, such as app store reviews and social media platforms. To address these challenges, in this dissertation, we incorporate methods of requirements modeling, data mining, domain engineering, and market analysis to develop a novel set of algorithms and tools for automatically classifying, synthesizing, and modeling the crowd's feedback in the mobile app market. Our analysis includes a set of empirical investigations and case studies, utilizing multiple large-scale datasets of mobile user data, in order to devise, calibrate, and validate our algorithms and tools. The main objective is to introduce a new form of crowd-driven software models that can be used by app developers to effectively identify and prioritize their end-users' concerns, develop apps to meet these concerns, and uncover optimized pathways of survival in the mobile app ecosystem.
Williams, Grant S., "Modeling Crowd Feedback in the Mobile App Market" (2019). LSU Doctoral Dissertations. 4839.