Huoran Li

Characterizing Smartphone Usage Patterns from Millions of Android Users

he prevalence of smart devices has promoted the popular- ity of mobile applications (a.k.a. apps) in recent years. A number of interesting and important questions remain unan- swered, such as why a user likes/dislikes an app, how an app becomes popular or eventually perishes, how a user selects apps to install and interacts with them, how frequently an app is used and how much traffic it generates, etc. This paper presents an empirical analysis of app usage behaviors collected from millions of users of Wandoujia, a leading An- droid app marketplace in China. The dataset covers two types of user behaviors of using over 0.2 million Android apps, including (1) app management activities (i.e., installa- tion, updating, and uninstallation) of over 0.8 million unique users and (2) app network traffic from over 2 million unique users. We explore multiple aspects of such behavior data and present interesting patterns of app usage. The results provide many useful implications to the developers, users, and disseminators of mobile apps.

PRADA: prioritizing android devices for apps by mining large-scale usage data

Selecting and prioritizing major device models are critical for mobile app developers to select testbeds and optimize resources such as marketing and quality-assurance resources. The heavily fragmented distribution of Android devices makes it challenging to select a few major device models out of thousands of models available on the market. Currently app developers usually rely on some reported or estimated general market share of device models. However, these estimates can be quite inaccurate, and more problematically, can be irrelevant to the particular app under consideration. To address this issue, we propose PRADA, the first approach to prioritizing Android device models for individual apps, based on mining large-scale usage data. PRADA adapts the concept of operational profiling (popularly used in software reliability engineering) for mobile apps – the usage of an app on a specific device model reflects the importance of that device model for the app. PRADA includes a collaborative filtering technique to predict the usage of an app on different device models, even if the app is entirely new (without its actual usage in the market yet), based on the usage data of a large collection of apps. We empirically demonstrate the effectiveness of PRADA over two popular app categories, i.e., Game and Media, covering over 3.86 million users and 14,000 device models collected through a leading Android management app in China.

Deriving User Preferences of Mobile Apps from Their Management Activities

App marketplaces host millions of mobile apps that are downloaded billions of times. Investigating how people manage mobile apps in their everyday lives creates a unique opportunity to understand the behavior and preferences of mobile device users, infer the quality of apps, and improve user experience. Existing literature provides very limited knowledge about app management activities, due to the lack of app usage data at scale. This article takes the initiative to analyze a very large app management log collected through a leading Android app marketplace. The dataset covers 5 months of detailed downloading, updating, and uninstallation activities, which involve 17 million anonymized users and 1 million apps. We present a surprising finding that the metrics commonly used to rank apps in app stores do not truly reflect the users’ real attitudes. We then identify behavioral patterns from the app management activities that more accurately indicate user preferences of an app even when no explicit rating is available. A systematic statistical analysis is designed to evaluate machine learning models that are trained to predict user preferences using these behavioral patterns, which features an inverse probability weighting method to correct the selection biases in the training process.

Through a Gender Lens: Learning Usage Patterns of Emojis from Large-Scale Android Users

Based on a large data set of emoji using behavior collected from smartphone users over the world, this paper investigates gender-specific usage of emojis. We present various interesting findings that evidence a considerable difference in emoji usage by female and male users. Such a difference is significant not just in a statistical sense; it is sufficient for a machine learning algorithm to accurately infer the gender of a user purely based on the emojis used in their messages. In real world scenarios where gender inference is a necessity, models based on emojis have unique advantages over existing models that are based on textual or contextual information. Emojis not only provide language-independent indicators, but also alleviate the risk of leaking private user information through the analysis of text and metadata.Emojis have gained incredible popularity in recent years and become a new ubiquitous language for Computer-Mediated Communication (CMC) by worldwide users. Various research efforts have been made to understand the behaviors of using emojis. Gender-specific study is always meaningful for HCI community, however, so far we know very little about whether and how much males and females vary in emoji usage. To bridge such a knowledge gap, this paper makes the first effort to explore the emoji usage through a gender lens. Our analysis is based on the largest data set to date, which covers 134,419 users from 183 countries, along with their over 401 million messages collected in three months. We conduct a multi-dimensional statistical analysis from various aspects of emoji usage, including the frequency, preferences, input patterns, public/private CMC-scenario patterns, temporal patterns, and sentiment patterns. The results demonstrate that emoji usage can significantly vary between males and females. Accordingly, we propose some implications that can raise useful insights to HCI community.

Understanding Diverse Usage Patterns from Large-Scale Appstore-Service Profiles

The prevalence of smart mobile devices has promoted the popularity of mobile applications (a.k.a. apps). Supporting mobility has become a promising trend in software engineering research. This article presents an empirical study of behavioral service profiles collected from millions of users whose devices are deployed with Wandoujia, a leading Android app-store service in China. The dataset of Wandoujia service profiles consists of two kinds of user behavioral data from using 0.28 million free Android apps, including (1) app management activities (i.e., downloading, updating, and uninstalling apps) from over 17 million unique users and (2) app network usage from over 6 million unique users. We explore multiple aspects of such behavioral data and present patterns of app usage. Based on the findings as well as derived knowledge, we also suggest some new open opportunities and challenges that can be explored by the research community, including app development, deployment, delivery, revenue, etc.

PRADO: Predicting App Adoption by Learning the Correlation between Developer-Controllable Properties and User Behaviors

To survive and stand out from the fierce market competition nowadays, it is critical for app developers to know (desirably ahead of time) whether, how well, and why their apps would be adopted by users. Ideally, the adoption of an app could be predicted by factors that can be controlled by app developers in the development process, and factors that app developers are able to take actions on and improve according to the predictions. To this end, this paper proposes PRADO, an approach to measuring various aspects of user adoption, including app download and installation, uninstallation, and user ratings. PRADO employs advanced machine learning algorithms to predict user adoption based on how these metrics correlate to a comprehensive taxonomy of 108 developer-controllable features of the app. To evaluate PRADO, we use 9,824 free apps along with their behavioral data from 12.57 million Android users, demonstrating that user adoption of a new app can be accurately predicted. We also derive insights on which factors are statistically significant to user adoption, and suggest what kinds of actions can be possibly performed by developers in practice.

Mining Device-Specific Apps Usage Patterns from Appstore Big Data

When smartphones, applications (a.k.a, apps), and app stores have been widely adopted by the billions, an interesting debate emerges: whether and to what extent do device models influence the behaviors of their users? The answer to this question is critical to almost every stakeholder in the smartphone app ecosystem, including app store operators, developers, end-users, and network providers. To approach this question, we collect a longitudinal data set of app usage through a leading Android app store in China, called Wandoujia. The data set covers the detailed behavioral profiles of 0.7 million (761,262) unique users who use 500 popular types of Android devices and about 0.2 million (228,144) apps, including their app management activities, daily network access time, and network traffic of apps. We present a comprehensive study on investigating how the choices of device models affect user behaviors such as the adoption of app stores, app selection and abandonment, data plan usage, online time length, the tendency to use paid/free apps, and the preferences to choosing competing apps. Some significant correlations between device models and app usage are derived from appstore big data, leading to important findings on the various user behaviors. For example, users owning different device models have a substantial diversity of selecting competing apps, and users owning lower-end devices spend more money to purchase apps and spend more time under cellular network.

Mining usage data from large-scale Android users: challenges and opportunities

Mining usage data from a large number of Android users can assist various software engineering tasks. In collaboration with Wandoujia, a leading Android app marketplace in China, we have conducted a large empirical analysis based on mining app usage behaviors collected from millions of Android users. Our empirical findings can provide implications, challenges, and opportunities to app-centric software development, deployment, and maintenance.