Sie sind auf Seite 1von 11

Attention and Engagement-Awareness in the Wild:

A Large-Scale Study with Adaptive Notifications


Tadashi Okoshi , Kota Tsubouchi , Masaya Taji , Takanori Ichikawa , and Hideyuki Tokuda
Graduate School of Media and Governance, Keio University
slash@ht.sfc.keio.ac.jp
Yahoo Japan Corporation

AbstractIn todays advancing ubiquitous computing age, in computing. Push notifications that pop up in the background
with its ever-increasing amount of information from various of a users attention at random times cause interruptions
applications and services available for consumption, the man- and divided attention. There have been several reports on
agement of peoples attention has become very important. In the negative effects caused by divided attention in terms of
particular, the high volume of notifications on mobile devices productivity, emotion, and mental state [2], [3], [4], [5].
has become a major cause of interruption of users. There has
been much research aimed at detecting the opportune moment Researchers have been investigating user interruptibility in
to present such information to users with in a way that lowers various ubiquitous and pervasive computing situations using
the cognitive load or frustration. However, evaluation of such different techniques with the objective of ensuring that inter-
systems in the real-world production environment with real users
and notifications, and evaluation on users engagement to the
ruptive notifications do not unnecessarily steal users precious
presented notification beyond simple responsiveness have not been attention resources. Breakpoint [6], the boundary between two
adequately studied. To the best of our knowledge, this study adjacent units of user activities, is known as a timing that
is the first to investigate user interruptibility and engagement can lower the impact on users cognitive load. We previously
using a real-world large-scale mobile application and real-world investigated the real-time detection of users breakpoints in
notifications consisting of actual news content. We equipped the their device interactions and physical activities using mobile
Yahoo! JAPAN Android app, one of the most popular applications sensing and machine learning techniques on smartphones [7]
on the national market, with our mobile-sensing and machine- and wearable watch devices [8].
learning-based interruptibility estimation logic. We conducted a
large-scale in-the-wild user study with more than 680,000 users However, we found that three significant issues remain
for three weeks. The results show that in most cases delaying the to be studied: (1) real-world evaluation of breakpoint-based
notification delivery until an interruptible moment is detected is adaptive notification with actual product application and as-
beneficial to users and results in significant reduction of user sociated notifications, (2) software architecture design of such
response time (49.7%) compared to delivering the notifications
interruptibility estimation for real-world deployment both on
immediately. We also observed a higher number of notifications
opened in our system as well as constant improvement in user the client and server sides, and (3) comprehensive evaluation
engagement levels throughout the entire study period. of user behavior in terms of not only interruptibility but also
users further engagement levels with the notification content.
I. I NTRODUCTION In this paper, we present the results and findings from a
While the capacity of our attention as humans is constant, large-scale study conducted on smartphone users interrupt-
the amount of information available for consumption has been ibility and user engagement with a popular real-world smart-
growing by several orders of magnitude. Concomitant with phone application. We designed and implemented our real-time
advances in computing and multitasking operating systems, breakpoint detection and notification scheduling mechanisms
more devices, and more applications and services, increasing inside the Yahoo! JAPAN Android application [9] (shown
volumes of notifications that proactively convey information in Figure 1), one of the most popular applications in the
to users are resulting in a greater number of interruptions.
Versatile applications and services in the cloud are being
developed and utilized in this ubiquitous computing age. These
software and services generate enormous amounts of various
types of information for users, such as big data analysis,
schedule reminders, messages from social media friends, the
weather forecast, breaking news, and status updates from
devices. Such information is delivered to users through devices
such as smartphones and other mobile devices, wearable
watches, and even through ambient devices embedded in a
users environment. For better timeliness and speediness, the
provision of such information has progressively become more
proactive, and it is often delivered through push notification
systems.
In this information-overload world, the constant and limited (Front screen (left), Weather radar (center), Notification (right))
capacity of human attention has become a new bottleneck [1] Fig. 1. Screenshots of Yahoo! JAPAN Android Application [9]
national application market. Considering several real-world and recognized by a user, some amount of his/her attention
requirements related to simplicity, scalability, and efficiency, with limited capacity [18], [19] is allocated to the information
our mechanism particularly focuses on the users physical- carried by the notification. This situation is called divided
activity breakpoints [8], relying on activity recognition APIs attention [20].
on the smartphone platform. Using mobile machine learning
Past studies have been revealing several types of negative
techniques, the detection mechanism embedded in the app
influence of interruptive notification, such as productivity [2],
detects the users breakpoints in real time and shows incoming
[3], [4], [5], [21], [22], emotional and social attribution [21],
notifications at such timings.
and psycho-physiological states [3]. Needs for computing sys-
Our large-scale user study, conducted over three weeks tems that can adapt their behavior to human users attentional
with a total of 687,840 users, revealed the efficiency of resources have been gradually recognized, with an increasing
our proposition. We found that, in most cases, notifications number of literatures particularly on sensing users attentional
delivered at delayed breakpoint timings improved the users states.
overall click timing (earlier). While the notification delivery
delay due to additional breakpoint detection (as opposed to the III. R ELATED W ORK
conventional deliver immediately style approach) is trivial, There are two main targets for sensing a humans current
once the notification is delivered, a significantly reduced user attentional state: the users current cognitive load and inter-
response time (49.7%) was observed in our approach. We also ruptibility.
observed a higher number of notifications opened in our system
as well as constant improvement in user engagement levels
A. Sensing Users Cognitive Load
throughout the entire study period.
The contribution of this paper is three-fold. First, we In cognitive psychology, the concept of cognitive load
present the design and implementation of our interruptibility is defined as the total amount of mental effort allocated to
detection mechanism on a large-scale real-world smartphone working memory. Several different approaches for measuring
platform. Second, we discuss our large-scale in-the-wild user this load have been proposed, including (a) subjective rating-
study on user interruptibility and engagement conducted using based methods, (b) task performance-based methods, and (c)
an actual product and associated notifications in a real-world physiological response-based methods.
situation. Finally, we evaluate our work in terms of not only Several studies on the subjective rating-based approach
interruptibility but also further user engagement with the have shown that the measurement of cognitive load through
presented notification contents. The remainder of this paper is post hoc self-reporting is a relatively reliable methodology for
organized as follows. Section II explains the interruption over- mental effort assessment [23]. The most widely used tool for
load problem. Section III discusses related work. Section IV assessing a users cognitive load is the NASA Task Load Index
clarifies our research goals. Section V specifies requirements (NASA-TLX) [24]. Although use of this method is widespread,
for our solution. Section VI presents the system design and the post hoc nature of the approach makes it difficult to apply
architecture of our system, AtteliaY. Section VII describes our to versatile ubiquitous computing systems where an assessment
initial model training study. Section VIII reports on our large- needs to be completed in real time.
scale in-the-wild user study conducted with 687,840 users for
The measurement of a users task performance is used
three weeks, in terms of our experimental design, methodology,
to objectively assess the users cognitive load during task
results, and analysis. Section IX discusses further research
execution. The users performance regarding their primary and
opportunities arising from the user study. Section X concludes
focal task is used in the primary task measurements, whereas
this paper.
secondary task measurements exploit the performance of a
II. I NTERRUPTION OVERLOAD secondary task (often asked to be) executed simultaneously
with the primary task [23]. In this methodology, the variation
Our current computing life suffers from interruption in reaction performance indicates variations in cognitive load.
overload caused by large numbers of notifications presented However, this methodology may not be feasible in ubiquitous
in inappropriate ways. Interruption overload is one class of computing situations where a user conducts multitasking with
a broader information overload problem discussed in the frequent task switching between multiple tasks, making it
literature [10], [11], [12]. More studies have recently been difficult to measure the response performances of the users
conducted in the context of interruptions and multitasking [13], various types of tasks using uniform measurement criteria.
[14], [15], [16], [17].
The psycho-physiological response-based method includes
The main source of interruption overload is notifications several different techniques, such as tracking of eye movement
from computer system entities such as local operating systems, and pupil size [25], [26], [27], [28], readings from electro-
messaging services connected to other users, and various cardiograms (ECG), galvanic skin response (GSR) [26], [29],
applications. The notification in computer systems was origi- [30], electroencephalogram (EEG) [29], [28], heart rate (HR),
nally designed to provide newly available information to users and HR variability (HRV) [31], [32], [28]. Haapalainen et
in a more timely and speedy manner (than polling by the al. [33] found that, in desktop computing, the combinational
user). Since typical notification systems deliver notifications use of an electrocardiogram and heat flux is the most accurate
immediately to users as soon as they are available, the users at classifying low and high levels of cognitive load. Although
end up facing numerous interruptive notifications from the this approach looks promising in terms of detecting users
background of their current tasks at random timings, regardless cognitive load in real time, the burden placed on the target
of their timing preference. When a notification is perceived users is not trivial.
B. Sensing Users Interruptibility platforms (e.g., Android and iOS) not having opened
their APIs to control notifications. Thus, past studies
Rather than sensing a users cognitive load to represent
mainly used a custom sample application and/or re-
that users internal mental status relatively directly, several
lated custom-made notifications prepared in an ad hoc
researchers have proposed detecting users interruptibility
manner for their research user study. Although some
from the viewpoint of the source of possible interruptions. This
studies [48], [49] focused on real-world smartphone
class of research can be categorized into two main groups: (a)
notifications, their main contributions pertained to
estimation of interruptibility at a certain timing period based
analysis of the current situations.
on a users context, and (b) detecting the users breakpoint [6].
2) The system design for such interruptibility detection
The breakpoint is the boundary between two adjacent actions
and notification adaptation in real situations (with the
of a person, and was found to be the timing when interrupting
real-world applications used by real users) has not
the user results in relatively lower frustration and cognitive
been adequately studied.
overhead [21], [34], [35].
3) User engagement for information content presented
Following early research in the desktop computing do- via notifications, beyond the users initial responses
main [36], [37], [38], [39], [40], more studies have recently to the notifications (such as response time or click
been conducted in the mobile field. Ho et al. used wireless on- rate), have not been adequately evaluated.
body accelerometers to trigger interruptions when users change
activities [41] and found that interruptions at these transition
times reduce user annoyance. The most recent studies have IV. R ESEARCH G OALS AND A PPROACHES
been on widespread mobile and smartphone environments. Considering the research issues outlined above, this study
Fischer et al. identified breakpoints after phone calls and aims to investigate smartphone users interruptibility against
text messages [42]. They found that users tend to be more notifications and further engagement against the notified con-
responsive to notifications after these activities than at other tent in systematically estimated opportune timings. Espe-
random times. Hofte et al. used an experience sampling cially, this research features such investigation in a real-
methodology to collect information on location, transit status, world environment with a real application, real users, and real
company, and activities in order to build a model of inter- notifications. To achieve these goals, we took the following
ruptibility [43], particularly for phone calls. Pejovic et al. approaches after careful discussion.
expanded the use of context to detect interruptible moments
on smartphones, including user activity, location, time of day, A1. Embedding interruptibility estimation logic into a
and emotional states [44]. Recent studies have even detected market-leading smartphone application: We added our in-
user boredom [45] as yet another opportune moment for terruptibility estimation logic into the Yahoo! JAPAN An-
notifications and engagement level[46] as a further indicator droid application [9]. Yahoo! JAPAN has been popular in the
of users response to received information. Japanese market since its launch in 1996, with a search engine
At the system level, current situation of fragmented inter- market share of 32% [50] (its share of Yahoo!s worldwide
ruptive notification delivery over mobile network is also known market share is 3.4%). The Android app, shown in Fig. 1,
to be inefficient in power consumption. Acer et al. [47] showed has an installed base of more than 10 million users, making it
that delaying notification delivery can yield power savings in one of the most popular smartphone applications. The app is a
mobile devices. portal-like application with several different features including
Web search, news reader, weather map, and links to a variety
Our previous works, Attelia I [7] and II [8], have followed of Yahoo! JAPAN services. To the best of our knowledge, we
the same research trend in interruptibility on smartphones are the first to conduct interruptibility research with a real-
and wearable watches (multi-device environment). Towards the world application that is utilized by such a large number of
realization of opportune moment detection and on-the-fly adap- users.
tation in notification scheduling, we particularly emphasized
four design principles: (1) feasibility on users real mobile A2. Using real-world notifications: Along with the applica-
and wearable devices, (2) supporting real-time detection, (3) tion, we use real notifications issued from Yahoo! services on
applicability to diverse types of applications, and (4) affinity the app to evaluate our interruptibility estimation approach.
to all-day use. Attelia realizes real-time breakpoint detection Whereas most previous studies used an artificial notification
on smartphones and wearable watches without any external or ESM [51] as notifications, utilization of real notifications
sensors or modification to existing operating systems or appli- from real information sources enables us to understand how
cations. users behave in real situations.
C. Further Research Challenges A3. Investigating users engagement: Finally, we quan-
titatively measure user engagement levels for the content
Although the studies cited above show that researchers are presented through each notification in addition to several
actively working on interruptibility and notification scheduling immediate response criteria, such as response time and click
on smartphones and wearable devices, there remain significant rate. Because users load Web content from Yahoo! JAPAN
research challenges that need to be addressed: servers when they click notifications, we decided to track
1) To the best of our knowledge, no study has inves- users browsing behavior from the server side by measuring
tigated and evaluated user interruptibility with real engagement-related criteria such as session length and revisit
(product-level) applications and the real notifications rate. This evaluation facilitates our understanding of how
issued from such applications. This is primarily ow- users behave and engage with the presented content beyond
ing to the current situation of major smartphone immediate click to open behavior.
V. R EQUIREMENTS FOR S YSTEM D ESIGN AND
D EVELOPMENT
In spite of the goals and approaches above, doing such
research on the real production environment has several dif-
ferences from that on a research-purpose environment. Here
we present the requirements for the system design and devel-
opment that emerged after our discussion in the beginning of
this project. Since we aim to port the original research software
Attelia into the Yahoo! JAPAN commercial production system,
we faced several real-world requirements related to acceptable
behavior and user experience of the product application, as
well as Yahoo!s business-oriented decision and restrictions.
R1. Users additional burden needs to be minimized.: When
placing the interruptibility estimation as an additional logic
to an existing application, depending on the sensor and API Fig. 2. Two Types of Breakpoints - Device Interaction and Physical Activity
types the logic uses, the users of the app will experience
additional burden, such as explicit confirmation to give ad-
ditional permissions (e.g., accessibility, location information) A. Detection of Physical Activity Breakpoints
to the application. Such burden should be minimized in order
to retain the existing user base of the application. For our concrete interruptibility estimation design to embed
inside the app, we decided to use and extend our previous
R2. Power overhead needs to be minimized.: As power is
work on real-time mobile detection of breakpoints developed
always a precious resource in mobile devices, and because
in Attelia [7], [8]. We previously placed breakpoints into
mobile users are very conscious about an applications power
two classes, namely physical activity breakpoints and device
usage, our system design needs to be energy-aware and mini-
interaction breakpoints as shown in Figure 2. In the figure,
mize power overhead.
a user is sitting down and doing work on her tablet. After a
R3. Cross-platform generalizability needs to be consid- while, she decides to take a coffee break. She stands up, walks
ered.: Although our first step experimental system can be to the kitchen, pours a coffee, walks back, sits down on the
implemented on a single platform for research purpose, the couch, and enjoys her coffee while watching a video on her
fundamental system design needs to be aware of cross-platform smartphone. In essence, in our daily lives with smartphones
generalizability over both iOS and Android major platforms. and wearable devices, there is a significant amount of time
when we simply carry or wear them but do not actively
R4. Collection of sensitive data needs to conform to use (manipulate) them, in contrast to the certain periods we
the corporate policy and process.: Additional collection actually do use (manipulate) them. By detecting two different
of sensitive and/or privacy-related sensor data, such as fine- types of breakpoints, our previous work detected interruptive
grained location information, needs to be proposed, carefully moments in users ubiquitous computing life comprehensively.
discussed, and approved in the corporate-wide business process
for assuring end users privacy protection. It means that this Meanwhile, in the present study, we pay special attention
process can take time and our system design may need to to the utilization of physical activity breakpoints. This is
start with a minimum set of data collection for the time-bound done for several reasons. Collecting UI interaction data on
research period. smartphone needs users explicit permission for accessibility
API on the smartphone platform (against R1). Furthermore,
R5. System design and development needs to conform accessing such sensitive information of users take long time
to the existing product management.: Yahoo! JAPAN in the product management process and it is even not clear if
Android application includes lots of commercial level prod- such data collection gets approved (against R4). On the other
uct features of lots of Yahoo! JAPAN services. Its product hand, detection of physical activity breakpoints has several
planning, design, and development processes are managed advantages. Physical activity breakpoints can solely cover a
in the business-oriented governance. Thus, the design and users all-day computing life as long as the user is carrying
development of our system naturally needs to fit such existing the device (even during the users active device use period).
processes. This also means that opportunity of the application Moreover, activity recognition API used for physical activity
updates to the market (i.e., GooglePlay) is considerably limited breakpoint detection has been recently implemented in both of
compared to single-purpose research prototype application the major mobile platforms [52], [53] (compatible with R3),
which often contains only bare-bone features and can be and those APIs are considered optimized in terms of efficiency,
pushed even nightly. accuracy, and power consumption (compatible with R2).

In our design, the system utilizes Google Play Services


VI. S YSTEM D ESIGN Location API (activity recognition API) [53] provided by the
Android platform, reads its detected activity results (e.g., walk-
In accordance with the goals, approaches, and the require- ing, still, on bike), and detects changes in the users activity
ments outlined above, this section explains our system design. as candidate timings of opportune moments for notification.
B. Notification to Be Tested 1) Notification Content Fetching: On the client side, News
Fetcher is a component that triggers whole breakpoint detec-
Table I gives the definitions of several different notification tion. News Fetcher maintains a connection to the server via
classes currently used in Yahoo! JAPAN services. For the first a TCP connection. When a new content (to be presented to
step of our study, we decided to use the Recommendation the user) is received from the server, News Fetcher notifies
class, for three reasons. First, this class of notifications consists the Controller component that initiates the core breakpoint
mainly of updates from services such as sports and showbiz- detection logic.
related news departments and does not include any emergency
content that needs to be presented to users immediately, so we 2) Life-cycle of Breakpoint Detection: Figure 4 shows the
can explore various notification scheduling techniques. Second, life-cycle of the core breakpoint detection logic. The core logic
receiver users of Recommendation notifications, i.e., those will be initiated upon the arrival of new notification content.
who enabled reception of this information class on the apps Once the logic starts, it repeatedly executes core detection
preference screen, receive exactly the same content, so we can logic (sensing, feature extraction, and prediction) every time
focus on investigating a users behavior and interruptibility a new sensor data is detected by Mobile Sensing. When a
without being influenced by any personalization aspect of the breakpoint is detected by Predictor (with an installed model),
notification content. Finally, this plan also fits the current the system issues a new Android notification and finishes the
product planning of the Yahoo! JAPAN application (compatible core detection logic.
with R5).
TABLE I. N OTIFICATION C LASSES

Class Name Real-Time/Batched Content Examples


Breaking news real-time breaking news articles
Natural disaster real-time weather update, earthquakes
Recommendation batched sports & showbiz news, tips
Personal batched transit, email, travel, auction

The recommendation notifications have four basic notifi-


cation timing slots (8AM, noon, 6PM, and 9PM) every day.
The content of the notifications are posted in the batch content
push queue by the news provider department in advance and
then sent to each user at the scheduled timings. At each timing
slot, only one notification content will be pushed to the users.
Fig. 4. Life-cycle of Breakpoint Detection Logic
C. System Architecture
With this design, we can minimize the duration of on-the-
Figure 3 shows the architecture of our production system. fly mobile sensing and online prediction and also minimize
Our implementation consists of a series of additional compo- the power consumption overhead (compatible with R2). We
nents inside the Yahoo! JAPAN Android application as well actually tested the additional power consumption in quality
as the components on the server side. assurance department and confirmed that the additional over-
head was less than 3% compared to the power consumption
of the original application.
3) Mobile Sensing: The Mobile Sensing component ob-
tains several types of sensor data, including that from the
GooglePlay Service Location API ActivityRecognition
and other device-related data. Table II lists the sensor data
collected on the client. The data mainly consists of a series
of output values from the GooglePlay Service Location APIs
ActivityRecognition and other device-based events (e.g.,
screen on/off events). On the activity recognition results from
ActivityRecognition, we used data only with a confidence
value greater than 51 (the value can be between 0 and 100)
based on our empirical knowledge. The mobile sensing compo-
nent obtains all of these data through individually implemented
event handlers for each sensor. Any changes in sensor data,
such as when the users activity changes or when the device
volume is changed, will be sensed and logged by the system.
4) Feature Extractor: When new sensor data is detected,
the Feature Extractor and Predictor modules execute and
predict if the current moment is a breakpoint of the user.
Fig. 3. System Architecture
Table III lists the features extracted from the sensor data. The
total number of features is 387. The system extracts seven
TABLE II. C OLLECTED S ENSOR DATA
7) Model Training and Update at The Server: The log data
Sensor Name Values along with the ground truth annotation will be sent to the
Activity Recognition IN VEHICLE, ON BICYCLE, ON FOOT, server nightly. At the server side, every night a new model
STILL, TILTING, UNKNOWN is built from all the data uploaded from the clients in the
Device Volume integer (0 to 7) past. The resulting models parameter will be available on the
Network Mode off, Mobile, Wi-Fi, WiMAX, LTE Model Server and will be downloaded to the clients on a daily
Screen Sleep normal, sleeping basis. This scheme of the periodic model update at the server
Vibration off, on and distribution of it to the clients nicely fits our requirement
R5, since the system can update the model frequently without
building and distributing the client application.
types of features: timestamp (hour), activity type, volume,
device sleep/awake status, vibration status, silent mode setting, Because the size of the client log is expected to be huge,
and network connection type. In addition to these sensor we prepared a dedicated Hadoop [54] cluster with 32 workers
values, transition of the sensor data is introduced as an eighth for model training. After the log data from the clients are
type of feature. For all possible From and To pairs of split into 32 buckets at the server side, each worker creates
sensor status transition, we prepared a dedicated feature value. a logistic linear regression on LIBLINEAR [55]. The 32
To detect breakpoints that occur in the timings of activity generated models will then be merged into a final unified
change (e.g., tilting their phone from the still status), model considering the bias of the number of training data.
these transition type features are helpful for the system to
characterize changes detected in each sensor. VII. I NITIAL M ODEL T RAINING
TABLE III. F EATURES E XTRACTED F ROM T HE S ENSOR DATA On the basis of the system design presented above, we
implemented a prototype version (a new version of the Ya-
Event Types Explanation Levels
hoo! JAPAN Android app) and conducted an initial study
Timestamp (Hour) Hour value extracted from time stamp of client 24 to collect users response logs to notifications and train the
Collected sensor data Present status of the client device 34 initial prediction model (that will be used in the production
Trigger Sensor type triggering this detection process 3 user study later), as well as to check the basic behavior
Transitions of From and To transition sets of each sensor 326 of our implementation. We installed the app on 39 Android
sensor values activity recognition (6x6), smartphone devices belonging to members of our lab and
volume(15x15), silent mode(4x4), network(6x6)
continued the study for 35 days.
We referenced the detailed specifications of the Android During the study, notifications from the real Yahoo! service
platform to specify the number of possible values in activ- were delivered at the conventional random timings. Users
ity transition, volume transition, silent mode transition, and response time to the notification along with the sensor data
network connection transition. We also confirmed that, to in the timing of the user click were logged on the client and
ensure the desired cross-platform nature of the system design, sent to the server. As described in Section VI-C6, the sensor
these feature designs are feasible and reasonable for future data will have annotated as a breakpoint ground truth when
implementation of the system on the iOS platform. the user clicks a notification within 10 seconds and vice versa.
5) Predictor: Using the above features, the actual break- The prediction models, trained from all the log and an-
point prediction is executed in Predictor with an installed notation data over 35 days, showed an average accuracy of
linear regression model. The model parameters are generated 94.8% in our 10-fold cross-validation. In this model, a total
on the server side and then downloaded to the clients. of 104 features (all those with a value over 0) are selected for
predicting breakpoints. Table IV shows the top 5 influencing
It is possible that no breakpoint is detected in Predictor for features. Interestingly, the top feature is the hour value of
a long period after the core detection module has been initiated. the clock being equal to 10. We speculate that this may be due
For example, when a user has placed the phone on the desk and to the working style of the participants (our lab members). The
no sensor data update (activity recognition, volume change, second to fifth features relate to the activity change occurrence
etc.) is detected, the core logic may not detect any breakpoints of different From and To activities. It is interesting to see
until a change in a time-related feature value influences a that IN VEHICLE (literally meaning that the user is in a
prediction result. In such cases, the system automatically issues vehicle) appears twice in this ranking. According to Googles
a notification after a specified timeout T. We set this value as API definition [53], this class means that the device is in a
1 hour throughout this research. vehicle but does not necessarily mean that the owner is driving
6) Ground Truth Annotation: The users reaction to the the vehicle. ON FOOT includes both walking and running.
notification, along with the current sensor data, will be logged
on the client. When a user clicks a notification within 10 TABLE IV. T OP 5 I NFLUENCING F EATURES IN THE I NITIAL M ODEL
seconds of the breakpoint (that triggered the notification),
all the sensor data at that breakpoint timing will be logged Feature Name Value

with an annotation of breakpoint ground truth. According to Hour value=10 4.27


Yahoo! JAPANs knowledge on notifications, empirically we Activity (STILL to ON FOOT) 3.51
configured the 10 second timeout value for this study. The Activity (STILL to IN VEHICLE) 3.50
log data is periodically sent to the server for further model Activity (IN VEHICLE to STILL) 3.30
Activity (TILTING to STILL) 3.23
training.
The API definition considers TILTING to be detected when Selection of the participating users was not visible to
the device angle relative to gravity changed significantly. This the participants. The application release was checked and
often occurs when a device is picked up from a desk or a user confirmed by the corporate legal and compliance departments.
who is sitting stands up. [53]. Although these five features do The study was conducted in conformity to the corporation
not solely determine a breakpoint, we were convinced enough regulation and the end user agreement.
by these results after a review discussion among the members
(including some subjective impressions). B. Experiment Procedure

VIII. E VALUATION The user study was conducted for three weeks (21 days)
in September 2016. To ensure the stability of the production
On the basis of these promising results from our initial application, the new version (including our implementation)
study, we conducted a large-scale in-the-wild user study was released to the production environment with a graduated
in the production environment with about 680,000 users for deployment scheme on the app store. After three days, the new
three weeks to better understand how our breakpoint-based version was made available for all users.
notification scheduling works in a real user environment. Our
evaluation criteria are as follows. The mechanism we utilized for this user study is the one
detailed in Section VI. Our logic was enabled only for the
1. Investigating users immediate response to the experimental group users and not enabled for the control group
breakpoint-scheduled notifications: We want to see how ef- users. Note that, except for the delivery timing difference of
fectively the breakpoint detection works from several different the specific Recommendation type notifications described in
points of view, such as relationships between activity change Section VI-B, the users in both groups experienced the same
and the conclusive detected breakpoints, and actual delay of notification content and delivery timings.
the notification by waiting for a breakpoint.
All users (of the experimental group) used the same model
2. Investigating users response to the breakpoint- for breakpoint prediction. At the beginning of the study, the
scheduled notifications: We want to observe how users react model trained in our initial study was installed in each clients
to the notifications scheduled at detected breakpoint timings. device. Once the study began, a new model was trained every
night at our Hadoop cluster from all clients log data and then
3. Investigating users (long-term) engagement to the pre- was downloaded to each client as a daily update.
sented contents and services: We want to investigate how
the users engagement level to the source Web service of the
C. Results and Analysis
notifications (Yahoo! JAPAN) will be influenced in the longer
term, beyond the users immediate reaction in a short period Through the nightly model update training over 21 days,
of time. our prediction model was gradually adapted to the users real
usage. After 21 days (i.e., 21 iterations of the model update),
A. Participants the latest model showed the average performance accuracy of
91.6% in the same 10-fold cross-validation methodology that
We selected 687,840 users (approximately 10% of the total we used in Section VII.
user base of the Yahoo! Android application) as participants in
this study. We used an existing A/B test infrastructure inside 1) Breakpoint and Activity Change: Table VI shows a
our application where a specific functional component of the breakdown list of detected breakpoints with true annotation
app can be enabled (or disabled) for a specific sub class of (i.e., breakpoints with a notification (presented based-on the
users based on the device ID. Using such system, we randomly basis of the breakpoint detection trigger) that was clicked by
selected 5% of the whole userbase as the experimental group the user within 10 seconds) into activity change pairs.
and the control group respectively. TABLE VI. B REAKDOWN OF T RUE - LABELED D ETECTED
Table V shows the demographics of the users. We split B REAKPOINTS INTO ACTIVITY C HANGES
these users evenly into two groups: (a) the experimental group To
(users to which our interruptibility detection and notification IN_VEHICLE ON_BICYCLE ON_FOOT STILL UNKNOWN TILTING
schedule are enabled), and (b) the control group (users to IN_VEHICLE 0.01 0.22 6.05 0.99 2.71
which our logic is not used) to compare the results and validate ON_BICYCLE 0.03 0.01 0.32 0.01 0.09
ON_FOOT 0.48 0.00 2.68 0.28 1.29
the effectiveness of our system. From
STILL 8.16 0.06 1.75 7.42 43.28
UNKNOWN 0.84 0.00 0.19 3.51 1.50
TABLE V. U SER D EMOGRAPHICS
TILTING 2.63 0.04 0.62 13.55 1.27
Number of Users 687,840
Gender Male 60.7% Very interestingly, the STILL to TILTING activity
Female 39.3% change showed the highest value. Again, as Googles API
0-19 3.6% document [53] mentions, timings such as when a device is
20-29 7.9% picked up from a desk or a user who is sitting stands up
30-39 22.3% are considered to have been opportune moments for the noti-
Age Group 40-49 35.2% fication receiver users. Moreover, activity changes to STILL
(Median: 44.0) 50-59 21.0% showed high numbers, such as TILTING to STILL (13.55%)
(Stdev.: 12.3) 60-69 8.4% and IN VEHICLE to STILL (6.05%). This matches our
70-79 1.4% previous hypothesis [8] that people would have breakpoints
80- 0.3%
when changing from a high energy state to a low energy state. control group is 3,258.1 seconds (standard deviation: 1,920.6).
On the other hand, we see that STILL to IN VEHICLE Comparing the two groups, we see that user response time was
resulted in the third highest number, 8.16%. What types of reduced by 49.7%, with statistical significance. Combined with
real-world situations are caught by this change is a possible the fact that more than 90% of notifications were delivered
topic for future research. (Figure 5), these result mean that, in most cases, the users
clicks of notifications occurred earlier in our breakpoint-
2) Notification Delay Due To Breakpoint: Figure 5 shows based notification scheduling. We conclude that, in most
a cumulative distribution function (CDF) on the delay from cases, delay in notification delivery due to the breakpoint
when a notification content arrives at the client to when a core detection does not hurt and is even beneficial because the users
estimation logic detects a breakpoint and the actual notification can click earlier.
is posted. The graph on the right side shows the overview. We
configured the timeout to 1 hour (3,600 seconds), so the value Figure 7 and 8 shows CDFs of two user groups on the
gets very close to 1 at 3,600 seconds. (We also observed a very response time for each of the four notifications issued every
rare situation where notification was further delayed due to an day. As explained previously, the Recommendation class
implementation issue.) When looking at the left side graph news content becomes available four times a day: 8AM, noon,
(zooming from 0 to 100 seconds), we see that more than 70% 6PM, and 9PM. When we plot each of them on the graph (blue:
of notifications were posted within approximately 10 seconds. 8AM, red: noon, green: 6PM, grey: 9PM), we confirm that,
The overall average delay is 236.8 seconds (approximately 4 for all of them, breakpoint-scheduled notifications resulted in
minutes). The timeout occurrence rate is 1.11%. clearly shorter response times. The average response times are,
in the experimental group, 1,490.0 sec (8AM), 1,527.8 sec
(noon), 1,824.2 sec (6PM), and 1,645.0 sec (9PM), and in the
control group, 3,306.9 sec (8AM), 2,995.7 sec (noon), 3,413.6
sec (6PM), and 3,400.6 sec (9PM).

Fig. 5. CDF on Delay from Content Reception To Notification Issue

3) Users Response to Notifications: One of the most


interesting results we obtained from this study is the users
response time to presented notifications. Figure 6 shows a
CDF on the users response time to the presented notifications
(from when a notification is posted to when a user clicks it)
in both user groups. The response times in the two groups are
significantly different: the time in the experimental group is
1,639.2 seconds (standard deviation: 1,690.2) while that in the

Fig. 7. CDF on Users Response Time To Notifications (Experimental Group)

Fig. 8. CDF on Users Response Time To Notifications (Control Group)


Fig. 6. CDF on Users Response Time To Notifications
4) Users Content Browsing: Table VII shows the number TABLE VIII. U SER E NGAGEMENT L EVEL IN B OTH G ROUPS
of opened notifications on a daily basis. Including the initial
three days when we gradually released the new application to
Timestamp Control Grp Experimental Grp gain (%)
GooglePlay, we measured the numbers for 24 days in total.
TABLE VII. U SER B ROWSING ON N OTIFICATION C ONTENT Before Study 3.757 3.758 0.030
Study Day 7 3.753 3.756 0.077
Day DOW Control Group Experimental Group gain(%)
Study Day 14 3.766 3.770 0.097
(Deploy Day 1) (Tue) 1,837 1,880 +2.34
Study Day 21 3.828 3.832 0.104
(Deploy Day 2) (Wed) 5,459 5,715 +4.69
(Deploy Day 3) (Thu) 9,096 9,280 +2.02 As seen in the table, the differences between the experi-
Study Day 1 (Fri) 29,837 30,386 +1.84 mental group and the control group continue to increase from
Study Day 2 (Sat) 27,747 27,918 +0.62 Before Study (0.030%) to Day 7 (0.077%), Day 14
Study Day 3 (Sun) 36,528 36,927 +1.09 (0.097%), and Day 21 (0.104%). We believe that this gain,
Study Day 4 (Mon) 39,990 41,253 +3.16 as much as 3.46 times in three weeks, is a promising result
Study Day 5 (Tue) 31,237 31,792 +1.78 in terms of long-term user engagement beyond simple Web
Study Day 6 (Wed) 33,869 34,320 +1.33 access session-wise user response to notifications.
Study Day 7 (Thu) 48,419 49,184 +1.58
Study Day 8 (Fri) 53,191 53,874 +1.28 IX. D ISCUSSION
Study Day 9 (Sat) 49,008 49,379 +0.76
Now we discuss the further research opportunities that this
Study Day 10 (Sun) 45,544 45,717 +0.38
research enables. First, we are interested in utilizing more
Study Day 11 (Mon)* 50,630 51,142 +1.01
types of sensors and features for exploring even better system
Study Day 12 (Tue) 39,136 41,266 +5.44
performance. In this study, the types of sensors were limited
Study Day 13 (Wed) 69,301 72,747 +4.97
due to time limitations related to both corporate approval
Study Day 14 (Thu)* 37,845 38,345 +1.32 and advance notification to the users. Once we have cleared
Study Day 15 (Fri) 42,394 42,971 +1.36 these requirements, injecting more types of sensor information
Study Day 16 (Sat) 47,824 48,051 +0.48 into the estimation model will be our future research focus,
Study Day 17 (Sun) 45,256 46,330 +2.37 while always maintaining our protection of users privacy. In
Study Day 18 (Mon) 33,128 33,678 +1.66 Figure 5, more than 70% of notifications were posted within
Study Day 19 (Tue) 39,089 39,479 +1.00 approximately 10 seconds. This can mean that, in some cases, a
Study Day 20 (Wed) (No data due to temporal system down) breakpoint was immediately detected at the first iteration of the
Study Day 21 (Thu) 42,569 43,159 +1.39 logic execution. We consider this comes from relative simple
(*): National public holiday in Japan set of features we used for the current system. For example, if
a trained model contains a relatively bigger weight for device
The interesting finding here is that, for 24 days in a row, volume set to value x, then the first execution immediately
the numbers in the control groups continued to get a greater detects a breakpoint for notification firing. Even so, we are
number of accesses. The average gain over the 21 days of the excited to our result in this study, with that limited amount of
user study period was 1.91% (standard deviation: 1.35). We sensor and feature types.
believe that this amount of gain is significant for Web services
Use of multiple models, including model personalization,
when compared with the industry practice. In addition, we
is another search avenue. In our user study, we observed an
can observe that we have more gain on weekdays than other
interesting difference in users responses between weekdays
days including weekends and holidays. The average gain of
and other days (Section VIII-C4). Further investigation on this
the weekdays is 2.39% (standard deviation: 1.45) whereas that
difference, as well as the difference between days (day-to-day
of weekends and holidays is 1.00% (standard deviation: 0.64).
difference) are obviously our immediate future work.
This difference may stem from differences of users behaviors
with regard to their physical activities (commuting to work vs. We are also interested in an even longer period of user
being at home) and also from net access activities on weekdays study in the production environment. During the present study,
and other days. we did not observe any statistical significance in differences
between user clicks and user engagement. However, we see
5) Longer-Term User Engagement: Finally, we analyzed
clearly reduced p values in some cases (on user engagement,
users longer-term engagement level to the Yahoo! JAPAN
p value decreased from 0.67 (before study) to 0.20 (day 21)).
Web service. As the evaluation criteria, we used a 5-level user
This motivates us to pursue further investigation over an even
engagement level that Yahoo! JAPAN typically uses for its
longer period.
user engagement analysis. This 5-level system was originally
designed to rank each user as 5 if she/he visits the Yahoo! Evaluation on the relationship between the notification
Web site 6-7 days a week, 4 in the case of 4-5 days, 3 in types (and content) and users preference and primary task
the case of 2-3 days, 2 in the case of 1 day, and 1 in the is also another future work. In this paper, for our first step,
case of 0 days. Introducing this metric, we summarized the we focused on the generic user response thus the evaluation.
score of 680,000 users on a group-by-group basis. Table VIII influence of notifications from other applications, including
shows the average level of users in each group. We first popular messenger applications, is yet to be studied. During
summarized the data a day before the study began and then our user study, we hypothesized that they are random external
did the same summary for days 7, 14, and 21 of the study. variables that can randomly occur in both user groups.
Finally, qualitative analysis from the end users perspective [13] Attention management in ubiquitous computing environments (amuce
is also among our future work. We could not do such evaluation 07), http://www.ubicomp.org/ubicomp2007/index-14.htm.
on the application itself during our user study, since our [14] S. Gould, D. Brumby, A. Cox, V. Gonzalez, D. Salvucci, and N. Taat-
experiment uses the real product application. A user interview gen, Multitasking and interruptions: a sig on bridging the gap between
research on the micro and macro worlds, in CHI12 Extended Abstracts
can be a concrete method of such evaluation. on Human Factors in Computing Systems, 2012, pp. 11891192.
[15] B. Poppinga, M. Pielot, N. Henze, N. Oliver, K. Church, and A. S.
X. C ONCLUSION Shirazi, Smarttention, please! intelligent attention management on
mobile devices, in Proceedings of the 17th International Conference
In todays advancing ubiquitous computing age, where on Human-Computer Interaction with Mobile Devices and Services
increasingly proactive notifications have been causing inter- Adjunct, ser. MobileHCI 15, 2015, pp. 10661069.
ruption overload, we addressed the problem of finding users [16] D. Weber, A. S. Shirazi, S. Gehring, N. Henze, B. Poppinga, M. Pielot,
interruptible moments in the mobile environment in order to and T. Okoshi, Smarttention, please!: 2nd workshop on intelligent
provide them with better notification experiences featuring attention management on mobile devices, in Proceedings of the 18th
International Conference on Human-Computer Interaction with Mobile
lower frustration, more responsiveness, and more engagement Devices and Services Adjunct, ser. MobileHCI 16, 2016, pp. 914917.
to the content. We developed a real-time interruptibility esti- [17] A. Voit, B. Poppinga, D. Weber, M. Bohmer, N. Henze, S. Gehring,
mation logic based on breakpoint detection inside the Yahoo! T. Okoshi, and V. Pejovic, Ubittention: Smart & ambient notification
JAPAN Android app, one of the most popular smartphone and attention management, in Adjunct Proceedings of the 2016 ACM
applications on the national market. Our large-scale in-the- International Joint Conference on Pervasive and Ubiquitous Computing,
wild user study in the production environment with more ser. UbiComp 16, 2016, pp. 15201523.
than 680,000 real users for 21 days clearly demonstrated the [18] D. Kahneman, Attention and effort. Prentice-Hall, Inc., 1973.
effectiveness of the system. We found that, in most cases, [19] R. M. Shiffrin and W. Schneider, Controlled and automatic human
notification delivery delay due to breakpoint detection does not information processing: Ii. perceptual learning, automatic attending and
a general theory. Psychological review, vol. 84, no. 2, p. 127, 1977.
hurt and even improves a users overall click timing (earlier),
with significantly reduced user response time (49.7%). We also [20] R. J. Sternberg and K. Sternberg, Cognitive Psychology, 6th ed. Cen-
gage Learning, 2012.
observed a continuous increase in content click numbers and
[21] P. D. Adamczyk and B. P. Bailey, If not now, when?: the effects of
user engagement level over the entire study period. interruption at different moments within task execution, in Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems,
R EFERENCES ser. CHI 04, 2004, pp. 271278.
[1] D. Garlan, D. Siewiorek, A. Smailagic, and P. Steenkiste, Project aura: [22] B. P. Bailey and J. A. Konstan, On the need for attention-aware
toward distraction-free pervasive computing, Pervasive Computing, systems: Measuring effects of interruption on task performance, error
IEEE, vol. 1, no. 2, pp. 22 31, april-june 2002. rate, and affective state, Computers in Human Behavior, vol. 22, no. 4,
pp. 685 708, 2006.
[2] J. G. Kreifeldt and M. E. McCarthy, Interruption as a test of the user-
computer interface, in JPL Proceeding of the 17 th Annual Conference [23] F. Paas, J. E. Tuovinen, H. Tabbers, and P. W. Van Gerven, Cognitive
on Manual Control, 1981, pp. 655667. load measurement as a means to advance cognitive load theory,
Educational psychologist, vol. 38, no. 1, pp. 6371, 2003.
[3] F. R. Zijlstra, R. A. Roe, A. B. Leonora, and I. Krediet, Temporal
factors in mental work: Effects of interrupted activities, Journal of [24] S. G. Hart and L. E. Staveland, Development of NASA-TLX (Task
Occupational and Organizational Psychology, vol. 72, no. 2, pp. 163 Load Index): Results of empirical and theoretical research, in Human
185, 1999. Mental Workload, ser. Advances in Psychology, P. A. Hancock and
N. Meshkati, Eds. North-Holland, 1988, vol. 52, pp. 139 183.
[4] C. Speier, J. S. Valacich, and I. Vessey, The influence of task
interruption on individual decision making: An information overload [25] J. Beatty and B. Lucero-Wagoner, The pupillary system, Handbook
perspective, Decision Sciences, vol. 30, no. 2, pp. 337360, 1999. of psychophysiology, vol. 2, pp. 142162, 2000.
[5] M. Czerwinski, E. Cutrell, and E. Horvitz, Instant messaging: Effects [26] C. S. Ikehara and M. E. Crosby, Assessing cognitive load with physi-
of relevance and timing, in People and computers XIV: Proceedings of ological sensors, in System Sciences, 2005. HICSS05. Proceedings of
HCI, vol. 2. British Computer Society, 2000, pp. 7176. the 38th Annual Hawaii International Conference on, 2005, pp. 295a
[6] D. Newtson and G. Engquist, The perceptual organization of ongoing 295a.
behavior, Journal of Experimental Social Psychology, vol. 12, no. 5, [27] S. T. Iqbal, P. D. Adamczyk, X. S. Zheng, and B. P. Bailey, Towards an
pp. 436450, 1976. index of opportunity: understanding changes in mental workload during
[7] T. Okoshi, J. Ramos, H. Nozaki, J. Nakazawa, A. K. Dey, and task execution, in Proceedings of the SIGCHI conference on Human
H. Tokuda, Attelia: Reducing users cognitive load due to interruptive factors in computing systems, 2005, pp. 311320.
notifications on smart phones, in Proceedings of IEEE International [28] G. F. Wilson, An analysis of mental workload in pilots during flight us-
Conference on Pervasive Computing and Communications 2015, ser. ing multiple psychophysiological measures, The International Journal
PerCom 15, 2015. of Aviation Psychology, vol. 12, no. 1, pp. 318, 2002.
[8] T. Okoshi, J. Ramos, H. Nozaki, J. Nakazawa, A. K. Dey, and [29] K. Ryu and R. Myung, Evaluation of mental workload with a combined
H. Tokuda, Reducing users perceived mental effort due to interruptive measure based on physiological indices during a dual task of tracking
notifications in multi-device mobile environments, in Proceedings and mental arithmetic, International Journal of Industrial Ergonomics,
of the 2015 ACM International Joint Conference on Pervasive and vol. 35, no. 11, pp. 9911009, 2005.
Ubiquitous Computing, ser. UbiComp 15, 2015, pp. 475486. [30] Y. Shi, N. Ruiz, R. Taib, E. Choi, and F. Chen, Galvanic skin response
[9] Yahoo! japan android application, https://play.google.com/store/apps/ (gsr) as an index of cognitive load, in CHI07 extended abstracts on
details?id=jp.co.yahoo.android.yjtop, Yahoo Japan Corporation, Sep. Human factors in computing systems, 2007, pp. 26512656.
2016.
[31] T. K. Fredericks, S. D. Choi, J. Hart, S. E. Butt, and A. Mital, An
[10] H. A. Simon, Designing organizations for an information-rich world, investigation of myocardial aerobic capacity as a measure of both
Computers, communication, and the public interest, vol. 37, pp. 4041, physical and cognitive workloads, International Journal of Industrial
1971. Ergonomics, vol. 35, no. 12, pp. 10971107, 2005.
[11] A. Toffler, Future shock. Bantam, 1990. [32] L. Mulder, Measurement and analysis methods of heart rate and
[12] J. T. Milord and R. P. Perry, A methodological study of overloadx, respiration for use in applied environments, Biological psychology,
The Journal of General Psychology, vol. 97, no. 1, pp. 131137, 1977. vol. 34, no. 2, pp. 205236, 1992.
[33] E. Haapalainen, S. Kim, J. F. Forlizzi, and A. K. Dey, Psycho- [50] StatCounter, Inc., StatCounter, http://statcounter.com/, Jul. 2016.
physiological measures for assessing cognitive load, in Proceedings [51] M. Csikszentmihalyi and R. Larson, Validity and reliability of the
of the 12th ACM international conference on Ubiquitous computing, experience-sampling method. The Journal of nervous and mental
ser. Ubicomp 10, 2010, pp. 301310. disease, vol. 175, no. 9, pp. 526536, 1987.
[34] S. T. Iqbal and B. P. Bailey, Investigating the effectiveness of mental [52] Apple Inc., CMMotionActivityManager, https://developer.
workload as a predictor of opportune moments for interruption, in CHI apple.com/library/ios/documentation/CoreMotion/Reference/
05 Extended Abstracts on Human Factors in Computing Systems, ser. CMMotionActivityManager class/index.html, 2014.
CHI EA 05, 2005, pp. 14891492.
[53] Google Inc., Making your app location-aware - Android Developers,
[35] S. T. Iqbal and B. Bailey, Leveraging characteristics of task structure https://developer.android.com/intl/ja/training/location/index.html.
to predict the cost of interruption, in Proceedings of the SIGCHI [54] The Apache Software Foundation, Apache Hadoop Project, http://
Conference on Human Factors in Computing Systems, ser. CHI 06, hadoop.apache.org/.
2006, pp. 741750.
[55] Machine Learning Group, National Taiwan University, LIBLINEAR
[36] E. Horvitz and J. Apacible, Learning and reasoning about interruption, A Library for Large Linear Classification, https://www.csie.ntu.edu.
in Proceedings of the 5th International Conference on Multimodal tw/cjlin/liblinear/.
Interfaces, ser. ICMI 03, 2003, pp. 2027.
[37] S. Hudson, J. Fogarty, C. Atkeson, D. Avrahami, J. Forlizzi, S. Kiesler,
J. Lee, and J. Yang, Predicting human interruptibility with sensors:
A wizard of oz feasibility study, in Proceedings of the SIGCHI
Conference on Human Factors in Computing Systems, ser. CHI 03,
2003, pp. 257264.
[38] J. B. Begole, N. E. Matsakis, and J. C. Tang, Lilsys: Sensing unavail-
ability, in Proceedings of the 2004 ACM Conference on Computer
Supported Cooperative Work, ser. CSCW 04, 2004, pp. 511514.
[39] E. Horvitz, P. Koch, and J. Apacible, Busybody: Creating and fielding
personalized models of the cost of interruption, in Proceedings of the
2004 ACM Conference on Computer Supported Cooperative Work, ser.
CSCW 04, 2004, pp. 507510.
[40] S. T. Iqbal and B. P. Bailey, Oasis: A framework for linking notifica-
tion delivery to the perceptual structure of goal-directed tasks, ACM
Transactions on Computer-Human Interaction, vol. 17, no. 4, pp. 15:1
15:28, Dec. 2010.
[41] J. Ho and S. S. Intille, Using context-aware computing to reduce the
perceived burden of interruptions from mobile devices, in Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems,
ser. CHI 05, 2005, pp. 909918.
[42] J. E. Fischer, C. Greenhalgh, and S. Benford, Investigating episodes
of mobile phone activity as indicators of opportune moments to deliver
notifications, in Proceedings of the 13th International Conference on
Human Computer Interaction with Mobile Devices and Services, ser.
MobileHCI 11, 2011, pp. 181190.
[43] G. H. H. ter Hofte, Xensible interruptions from your mobile phone, in
Proceedings of the 9th International Conference on Human Computer
Interaction with Mobile Devices and Services, ser. MobileHCI 07,
2007, pp. 178181.
[44] V. Pejovic and M. Musolesi, InterruptMe : Designing Intelligent
Prompting Mechanisms for Pervasive Applications, in Proceedings
of the 2014 ACM International Joint Conference on Pervasive and
Ubiquitous Computing, ser. UbiComp 14, 2014, pp. 395906.
[45] M. Pielot, T. Dingler, J. S. Pedro, and N. Oliver, When attention is not
scarce - detecting boredom from mobile phone usage, in Proceedings
of the 2015 ACM International Joint Conference on Pervasive and
Ubiquitous Computing, ser. UbiComp 15, 2015, pp. 825836.
[46] A. Mathur, N. D. Lane, and F. Kawsar, Engagement-aware computing:
Modelling user engagement from mobile contexts, in Proceedings
of the 2016 ACM International Joint Conference on Pervasive and
Ubiquitous Computing, ser. UbiComp 16, 2016, pp. 622633.
[47] U. Acer, A. Mashhadi, C. Forlivesi, and F. Kawsar, Energy efficient
scheduling for mobile push notifications, in Proceedings of the 12th
EAI International Conference on Mobile and Ubiquitous Systems:
Computing, Networking and Services on 12th EAI International Confer-
ence on Mobile and Ubiquitous Systems: Computing, Networking and
Services, ser. MOBIQUITOUS'15, 2015, pp. 100109.
[48] A. Sahami Shirazi, N. Henze, T. Dingler, M. Pielot, D. Weber, and
A. Schmidt, Large-scale assessment of mobile notifications, in Pro-
ceedings of the 32nd annual ACM conference on Human factors in
computing systems - CHI 14, 2014, pp. 30553064.
[49] M. Pielot, K. Church, and R. de Oliveira, An In-Situ Study of
Mobile Phone Notifications, in Proceedings of the 16th International
Conference on Human-Computer Interaction with Mobile Devices and
Services - MobileHCI 14, 2014, pp. 233242.

Das könnte Ihnen auch gefallen