Remo Manuel Frey

Understanding the impact of personality traits on mobile app adoption - Insights from a large-scale field study

The sheer amount of available apps allows users to customize smartphones to match their personality and interests. As one of the first large-scale studies, the impact of personality traits on mobile app adoption was examined through an empirical study involving 2043 Android users. A mobile app was developed to assess each smartphone users personality traits based on a state-of-the-art Big Five questionnaire and to collect information about her installed apps. The contributions of this work are two-fold. First, it confirms that personality traits have significant impact on the adoption of different types of mobile apps. Second, a machine-learning model is developed to automatically determine a users personality based on her installed apps. The predictive model is implemented in a prototype app and shows a 65% higher precision than a random guess. Additionally, the model can be deployed in a non-intrusive, low privacy-concern, and highly scalable manner as part of any mobile app. Personality has a significant impact on mobile app adoption.A novel approach is proposed to study mobile app adoption on a large scale.A machine-learning model is developed to predict a smartphone users personality.The predictive model can be integrated into any mobile app.

Mobile app adoption in different life stages: An empirical analysis

Abstract The analysis of individuals’ current life stages is a powerful approach for identifying und understanding patterns of human behavior. Different stages imply different preferences and consumer demands. Thus, life stages play an important role in marketing, economics, and sociology. However, such information is difficult to be obtained especially in the digital world. This work thus contributed to both theory and practice from two aspects. First, we conducted a large-scale empirical study with 1435 participants and showed that a person’s mobile app adoption pattern is strongly influenced by her current life stage. Second, we presented a data-driven, highly-scalable, and real-time approach of predicting an individual’s current life stage based on the apps she has installed on smartphone. Result showed that our predictive models were able to predict life stages with 241.0% higher precision and 148.2% higher recall than a random guess on average.

Mass Personalization through Mobile App Adoption Analytics

Reality mining on mobile devices is a young but emergent research field. Recent research articles propose to gather data about installed apps on smartphones as a way to get valuable user information and to derive user profiles. The kind of apps the users install and use are obviously linked to their interests, their demographics, and their personalities. They constantly install, update, or delete apps on their mobile devices to match the personal needs – thus making this data stream valuable for real-time user profiling. Especially because apps are favored over a web browser on mobile phones, and thus classical methods (e.g. cookies) fail. Using these profiles in business applications enables so-called “mass personalization”: Companies are able to personalize products, services, recommendations, prices, and digital content to all their customers in a fully automated manner. Timeconsuming and intrusive questionnaires for a better understanding of consumers’ needs and habits are no longer required. Thus, collecting a snapshot of users’ installed apps and analyzing it – hereinafter referred to as “Mobile App Adoption Analytics” – may bring new marketing and business opportunities. The aim of this dissertation is to examine the described app-based approach in detail, to gain scientific insights, and to provide the basis for practical applications. Here, the approach is viewed from three different perspectives: people’s app adoption behavior, potential applications, and privacy considerations. From the first perspective, the present work complements the findings by Runhua Xu from 2016 on mobile app adoption with the analysis of the adoption in different “life stages” (Chapter 3) and for different “life events” (Chapter 4). These two concepts are often described as major forces that are going to shape today’s and tomorrow’s consumer need and behavior. The conducted user studies are the first to reveal the changes across the different life stages. Typical behavioral patterns are consistent with app adoption behavior. For example, the diminished mobility behavior of young families is reflected in significantly fewer transportation and travel apps on their smartphones. The derived models for life stage prediction provide acceptable accuracy and precision. Similarly good results are provided by the models for life event prediction. In particular, the adoption behavior is outstanding in estimating whether a user expects to have a first child within six months, with a 50% precision. The second perspective considers potential applications. Consumer applications like recommender systems are needed to give the method a relevance in practice. Three applications are presented

Mobile recommendations based on interest prediction from consumer's installed apps–insights from a large-scale field study

Abstract Recommender systems are essential in mobile commerce to benefit both companies and individuals by offering highly personalized products and services. One key pre-requirement of applying such systems is to gain decent knowledge about each individual consumer through user profiling. However, most existing profiling approaches on mobile suffer problems such as non-real-time, intrusive, cold-start, and non-scalable, which prevents them from being adopted in reality. To tackle the problems, this work developed real-time machine-learning models to predict user profiles of smartphone users from openly accessible data, i.e. app installation logs. Results from a study with 904 participants showed that the models are able to predict interests on average 48.81% better than a random guess in terms of precision and 13.80% better in terms of recall. Since the effectiveness of such predictive models is unknown in practice, the predictive models were evaluated in a large-scale field experiment with 73,244 participants. Results showed that by leveraging our models, personalized mobile recommendations can be enabled and the corresponding click-through-rate can be improved by up to 228.30%. Supplementary information, study data, and software can be found at https://www.autoidlabs.ch/mobile-analytics .

Universal Food Allergy Number

In 2007 the European Union defined a list of 14 food ingredients which are likely to cause adverse reactions in susceptible individuals. As such, legislation mandates that these ingredients must be indicated on the label of relevant foodstuffs. However, there is no machine readable standard for the declaration of these ingredients. We propose to encode this information in a 5-digit number. The number can either be added to items on a menu card or printed as a barcode on food products. Further, we propose a complementary, 5-digit number which contains information about food allergies of an individual. The number is short enough to share verbally, for instance over a phone call for a restaurant reservation. By comparing both sets of numbers as a food allergy test, individual intolerances are immediately visible. As a proof of concept, we developed an app enabling users to quickly check whether or not foodstuffs are safe to consume based on their allergies.