Tapio Soikkeli

Diversity and End User Context in Smartphone Usage Sessions

Mobile end user context has gained increasing attention in the mobile services industry. Context information is seen as an important component in developing new, more personalized, mobile services and applications. This paper studies the effect of end user context on smart phone usage sessions. Smart phone usage sessions are used to depict user behavior and usage habits of smart phone users on a high level. We have detected end user contexts, and extracted smart phone usage session information from handset-based data of 140 smart phone users. We first examine and describe usage sessions as such, and then in different end user contexts. According to our usage session analysis, smart phone usage is highly diversified across users. For example, the average number of sessions per day ranges from 3 to 46. Characteristics of smart phone usage sessions differ in different end user contexts. For example, an average session is 37 % longer in the Home-context than in the Office-context, but Office has 56 % more sessions per time unit than Home. The results imply that mobile services and applications need to adapt to user behavior in order to be personalized enough, and that context awareness is indeed a worthwhile step towards this.

Contextual usage patterns in smartphone communication services

The mobile end user context has received a lot of attention from the mobile services industry lately. The location-based and context-sensitive information that are characteristic for smartphones can be utilized to study the use context of mobile end users. Accordingly, this article utilizes handset-based data in analyzing how the context of use affects the usage of smartphone communication services. The context is identified with an algorithm utilizing mobile network cell ID and WLAN data and resulting in five place-related contexts, namely Home, Office, Other meaningful, Elsewhere and Abroad. According to our analysis, voice calls are used least intensively in the Home context where the length of the voice calls is the longest, however. Email and SMS are used most intensively in the Office context, where the voice calls are the shortest in duration. Finally, mobile IM/VoIP and social media services are more free-time oriented as they are used most intensively in Elsewhere and Other meaningful contexts. The findings imply that people use smartphone communication services differently depending on the use context. However, context can be defined and identified in a number of ways, and this article presents only one solution that is highly dependent on the type of data collected.

Characterizing Smartphone Usage: Diversity and End User Context

Mobile end user context has gained increasing attention in the mobile services industry. This article utilizes handset-based data, collected from 140 users, to examine smartphone usage in different place-related end user contexts. Smartphone usage is examined first on a high level by using smartphone usage session as a unit of analysis. Then the usage sessions are dismantled into application sessions for deeper analysis and application level study. According to the authors’ analysis, smartphone usage is highly diversified across users. For example, the daily smartphone usage time differs by orders of magnitude between users. They observed also that smartphones are used differently in different end user contexts. For example, some applications are clearly more context sensitive than others. The results imply that mobile services and applications need to adapt to user behavior in order to be personalized enough, and that context awareness can indeed be a worthwhile step towards this.

Multidevice mobile sessions: A first look

Abstract The increasing number of users with multiple mobile devices underscores the importance of understanding how users interact, often simultaneously, with these multiple devices. However, most device based monitoring studies have focused only on a single device type. In contrast, we study the multidevice usage of a US-based panel through device based monitoring on panelist’s smartphone and tablet devices. We present a broad range of results from characterizing individual multidevice sessions to estimating device usage substitution. For example, we find that for panelists, 50% of all device interaction time can be considered multidevice usage.

Comparison of context-aware predictive modeling approaches: Semantic place in inferring mobile user behavior

Purpose – The aim of this paper is to empirically examine how to best incorporate such contextual data, such as location or the semantic place of mobile users, into mobile user behavior models. Acquiring such data has become technically easier than ever. The proper utilization of these data leads, hypothetically, to better understanding of mobile user behavior and, consequently, to enhanced mobile services. Design/methodology/approach – The paper systematically compares, under multiple experimental settings, the predictive performances of models built with three different approaches (pre-filtering, contextual modeling and post-filtering) used for incorporating contextual data into user behavior models. The comparisons focus on by which approach additional semantic place information can be best utilized for making the most accurate inferences on mobile user behavior. Real-life smartphone usage data are utilized in the analysis. Findings – The results demonstrate that none of the considered approaches unifo...

Mobile Device Type Substitution

Mobile users today interact with a variety of mobile device types including smartphones, tablets, smartwatches, and others. However research on mobile device type substitution has been limited in several respects including a lack of detailed and robust analyses. Therefore, in this work we study mobile device type substitution through analysis of multidevice usage data from a large US-based user panel. Specifically, we use regression analysis over paired user groups to test five device type substitution hypotheses. We find that both tablets and PCs are partial substitutes for smartphones with tablet and PC ownership decreasing smartphone usage by about 12.5 and 13 hours/month respectively. Additionally, we find that tablets and PCs also prompt about 20 and 57 hours/month respectively of additional (non-substituted) usage. We also illustrate significant inter-user diversity in substituted and additional usage. Overall, our results can help in understanding the relative positioning of different mobile device types and in parameterizing higher level mobile ecosystem models.