| Planning, Apps, and the High-End Smartphone: Exploring the Landscape of Modern Cross-Device Reaccess | | BIBA | Full-Text | 1-18 | |
| Elizabeth Bales; Timothy Sohn; Vidya Setlur | |||
| The rapid growth of mobile devices has made it challenging for users to maintain a consistent digital history among all their personal devices. Even with a variety of cloud computing solutions, users continue to redo web searches and reaccess web content that they already interacted with on another device. This paper presents insights into the cross-device reaccess habits of 15 smartphone users. We studied how they reaccessed content between their computer and smartphone through a combination of data logging, a screenshot-based diary study, and user interviews. From 1276 cross-device reaccess events we found that users reaccess content between their phone and computer with comparable frequency, and that users rarely planned ahead for their reaccess needs. Based on our findings, we present opportunities for building future mobile systems to support the unplanned activities and content reaccess needs of mobile users. | |||
| Understanding Human-Smartphone Concerns: A Study of Battery Life | | BIBAK | Full-Text | 19-33 | |
| Denzil Ferreira; Anind K. Dey; Vassilis Kostakos | |||
| This paper presents a large, 4-week study of more than 4000 people to assess
their smartphone charging habits to identify timeslots suitable for
opportunistic data uploading and power intensive operations on such devices, as
well as opportunities to provide interventions to support better charging
behavior. The paper provides an overview of our study and how it was conducted
using an online appstore as a software deployment mechanism, and what battery
information was collected. We then describe how people charge their
smartphones, the implications on battery life and energy usage, and discuss how
to improve users' experience with battery life. Keywords: Large-scale study; battery life; autonomous logging; smartphones; android | |||
| Monitoring Residential Noise for Prospective Home Owners and Renters | | BIBAK | Full-Text | 34-49 | |
| Thomas Zimmerman; Christine Robson | |||
| Residential noise is a leading cause of neighborhood dissatisfaction but is
difficult to quantify for it varies in intensity and spectra over time. We have
developed a noise model and data representation techniques that prospective
homeowners and renters can use to provide quantitative and qualitative answers
to the question, "is this a quiet neighborhood?" Residential noise is modeled
as an ambient background punctuated by transient events. The quantitative noise
model extracts noise features that are sent as SMS text messages. A device that
implements the noise model has been build, calibrated and verified. The
qualitative impact of sound is subjectively assessed by providing one-minute
audio summaries composed of twenty 3-second sound segments that represent the
loudest noise events occurring in a 24 hour sampling period. The usefulness and
desirability of the noise pollution monitoring service is confirmed with pre-
and post-use surveys. Keywords: Location-based services; mobile devices; sensors | |||
| A Longitudinal Study of Pressure Sensing to Infer Real-World Water Usage Events in the Home | | BIBAK | Full-Text | 50-69 | |
| Jon Froehlich; Eric C. Larson; Elliot Saba; Tim Campbell; Les Atlas; James Fogarty; Shwetak Patel | |||
| We present the first longitudinal study of pressure sensing to infer
real-world water usage events in the home (e.g., dishwasher, upstairs bathroom
sink, downstairs toilet). In order to study the pressure-based approach out in
the wild, we deployed a ground truth sensor network for five weeks in three
homes and two apartments that directly monitored valve-level water usage by
fixtures and appliances. We use this data to, first, demonstrate the practical
challenges in constructing water usage activity inference algorithms and,
second, to inform the design of a new probabilistic-based classification
approach. Inspired by algorithms in speech recognition, our novel Bayesian
approach incorporates template matching, a language model, grammar, and prior
probabilities. We show that with a single pressure sensor, our probabilistic
algorithm can classify real-world water usage at the fixture level with 90%
accuracy and at the fixture-category level with 96% accuracy. With two pressure
sensors, these accuracies increase to 94% and 98%. Finally, we show how our new
approach can be trained with fewer examples than a strict template-matching
approach alone. Keywords: Water sensing; activity inference; sustainability; field deployments | |||
| Exploring the Design Space for Situated Glyphs to Support Dynamic Work Environments | | BIBA | Full-Text | 70-78 | |
| Fahim Kawsar; Jo Vermeulen; Kevin Smith; Kris Luyten; Gerd Kortuem | |||
| This note offers a reflection on the design space for a situated glyph -- a single, adaptive and multivariate graphical unit that provides in-situ task information in demanding work environments. Rather than presenting a concrete solution, our objective is to map out the broad design space to foster further exploration. The analysis of this design space in the context of dynamic work environments covers i) information affinity -- the type of information can be presented with situated glyphs, ii) representation density -- the medium and fidelity of information presentation, iii) spatial distribution -- distribution granularity and placement alternatives for situated glyphs, and finally iv) temporal distribution -- the timing of information provision through glyphs. Our analysis has uncovered new problem spaces that are still unexplored and could motivate further work in the field. | |||
| Learning Time-Based Presence Probabilities | | BIBAK | Full-Text | 79-96 | |
| John Krumm; A. J. Bernheim Brush | |||
| Many potential pervasive computing applications could use predictions of
when a person will be at a certain place. Using a survey and GPS data from 34
participants in 11 households, we develop and test algorithms for predicting
when a person will be at home or away. We show that our participants'
self-reported home/away schedules are not very accurate, and we introduce a
probabilistic home/away schedule computed from observed GPS data. The
computation includes smoothing and a soft schedule template. We show how the
probabilistic schedule outperforms both the self-reported schedule and an
algorithm based on driving time. We also show how to combine our algorithm with
the best part of the drive time algorithm for a slight boost in performance. Keywords: Location prediction; presence prediction; away prediction; energy
efficiency; human routines | |||
| n-Gram Geo-trace Modeling | | BIBA | Full-Text | 97-114 | |
| Senaka Buthpitiya; Ying Zhang; Anind K. Dey; Martin L. Griss | |||
| As location-sensing smart phones and location-based services gain mainstream popularity, there is increased interest in developing techniques that can detect anomalous activities. Anomaly detection capabilities can be used in theft detection, remote elder-care monitoring systems, and many other applications. In this paper we present an n-gram based model for modeling a user's mobility patterns. Under the Markovian assumption that a user's location at time t depends only on the last n-1 locations until t-1, we can model a user's idiosyncratic location patterns through a collection of n-gram geo-labels, each with estimated probabilities. We present extensive evaluations of the n-gram model conducted on real-world data, compare it with the previous approaches of using T-Patterns and Markovian models, and show that for anomaly detection the n-gram model outperforms existing work by approximately 10%. We also show that the model can use a hierarchical location partitioning system that is able to obscure a user's exact location, to protect privacy, while still allowing applications to utilize the obscured location data for modeling anomalies effectively. | |||
| Autonomous Construction of a WiFi Access Point Map Using Multidimensional Scaling | | BIBAK | Full-Text | 115-132 | |
| Jahyoung Koo; Hojung Cha | |||
| To construct a WiFi positioning system, dedicated individuals usually gather
radio scans with ground truth data. This laborious operation limits the
widespread use of WiFi-based locating system. Off-the-shelf smartphones have
the capability to scan radio signals from WiFi Access Points (APs). In this
paper we propose a scheme to construct a map of WiFi AP positions autonomously
without ground truth information. From radio scans, we extract dissimilarities
between pairs of WiFi APs, then analyze the dissimilarities to produce a
geometric configuration of WiFi APs based on a multidimensional scaling
technique. To validate our scheme, we conducted experiments on five floors of
an office building that has an area of 50 m by 35 m in each floor. WiFi APs
were located within a 10m error range, and floors of APs are recognized without
error. Keywords: WiFi Access Point Map; Positioning; Autonomous and Unsupervised Learning;
Multidimensional Scaling | |||
| Identifying Important Places in People's Lives from Cellular Network Data | | BIBA | Full-Text | 133-151 | |
| Sibren Isaacman; Richard A. Becker; Ramón Cáceres; Stephen G. Kobourov; Margaret Martonosi; James Rowland; Alexander Varshavsky | |||
| People spend most of their time at a few key locations, such as home and work. Being able to identify how the movements of people cluster around these "important places" is crucial for a range of technology and policy decisions in areas such as telecommunications and transportation infrastructure deployment. In this paper, we propose new techniques based on clustering and regression for analyzing anonymized cellular network data to identify generally important locations, and to discern semantically meaningful locations such as home and work. Starting with temporally sparse and spatially coarse location information, we propose a new algorithm to identify important locations. We test this algorithm on arbitrary cellphone users, including those with low call rates, and find that we are within 3 miles of ground truth for 88% of volunteer users. Further, after locating home and work, we achieve commute distance estimates that are within 1 mile of equivalent estimates derived from government census data. Finally, we perform carbon footprint analyses on hundreds of thousands of anonymous users as an example of how our data and algorithms can form an accurate and efficient underpinning for policy and infrastructure studies. | |||
| NextPlace: A Spatio-temporal Prediction Framework for Pervasive Systems | | BIBA | Full-Text | 152-169 | |
| Salvatore Scellato; Mirco Musolesi; Cecilia Mascolo; Vito Latora; Andrew T. Campbell | |||
| Accurate and fine-grained prediction of future user location and
geographical profile has interesting and promising applications including
targeted content service, advertisement dissemination for mobile users, and
recreational social networking tools for smart-phones. Existing techniques
based on linear and probabilistic models are not able to provide accurate
prediction of the location patterns from a spatio-temporal perspective,
especially for long-term estimation. More specifically, they are able to only
forecast the next location of a user, but not his/her arrival time and
residence time, i.e., the interval of time spent in that location. Moreover,
these techniques are often based on prediction models that are not able to
extend predictions further in the future.
In this paper we present NextPlace, a novel approach to location prediction based on nonlinear time series analysis of the arrival and residence times of users in relevant places. NextPlace focuses on the predictability of single users when they visit their most important places, rather than on the transitions between different locations. We report about our evaluation using four different datasets and we compare our forecasting results to those obtained by means of the prediction techniques proposed in the literature. We show how we achieve higher performance compared to other predictors and also more stability over time, with an overall prediction precision of up to 90% and a performance increment of at least 50% with respect to the state of the art. | |||
| Using Decision-Theoretic Experience Sampling to Build Personalized Mobile Phone Interruption Models | | BIBAK | Full-Text | 170-187 | |
| Stephanie Rosenthal; Anind K. Dey; Manuela M. Veloso | |||
| We contribute a method for approximating users' interruptibility costs to
use for experience sampling and validate the method in an application that
learns when to automatically turn off and on the phone volume to avoid
embarrassing phone interruptions. We demonstrate that users have varying costs
associated with interruptions which indicates the need for personalized cost
approximations. We compare different experience sampling techniques to learn
users' volume preferences and show those that ask when our cost approximation
is low reduce the number of embarrassing interruptions and result in more
accurate volume classifiers when deployed for long-term use. Keywords: interruptibility; preference elicitation; mobile devices; machine learning | |||
| SpeakerSense: Energy Efficient Unobtrusive Speaker Identification on Mobile Phones | | BIBAK | Full-Text | 188-205 | |
| Hong Lu; A. J. Bernheim Brush; Bodhi Priyantha; Amy K. Karlson; Jie Liu | |||
| Automatically identifying the person you are talking with using continuous
audio sensing has the potential to enable many pervasive computing applications
from memory assistance to annotating life logging data. However, a number of
challenges, including energy efficiency and training data acquisition, must be
addressed before unobtrusive audio sensing is practical on mobile devices. We
built SpeakerSense, a speaker identification prototype that uses a
heterogeneous multi-processor hardware architecture that splits computation
between a low power processor and the phone's application processor to enable
continuous background sensing with minimal power requirements. Using
SpeakerSense, we benchmarked several system parameters (sampling rate, GMM
complexity, smoothing window size, and amount of training data needed) to
identify thresholds that balance computation cost with performance. We also
investigated channel compensation methods that make it feasible to acquire
training data from phone calls and an automatic segmentation method for
training speaker models based on one-to-one conversations. Keywords: Continuous audio sensing; mobile phones; speaker identification; energy
efficiency; heterogeneous multi-processor hardware | |||
| Text Text Revolution: A Game That Improves Text Entry on Mobile Touchscreen Keyboards | | BIBAK | Full-Text | 206-213 | |
| Dmitry Rudchenko; Tim Paek; Eric Badger | |||
| Mobile devices often utilize touchscreen keyboards for text input. However,
due to the lack of tactile feedback and generally small key sizes, users often
produce typing errors. Key-target resizing, which dynamically adjusts the
underlying target areas of the keys based on their probabilities, can
significantly reduce errors, but requires training data in the form of touch
points for intended keys. In this paper, we introduce Text Text Revolution
(TTR), a game that helps users improve their typing experience on mobile
touchscreen keyboards in three ways: first, by providing targeting practice,
second, by highlighting areas for improvement, and third, by generating ideal
training data for key-target resizing as a side effect of playing the game. In
a user study, participants who played 20 rounds of TTR not only improved in
accuracy over time, but also generated useful data for key-target resizing. To
demonstrate usefulness, we trained key-target resizing on touch points
collected from the first 10 rounds, and simulated how participants would have
performed had personalized key-target resizing been used in the second 10
rounds. Key-target resizing reduced errors by 21.4%. Keywords: Game; key-target resizing; text entry; touchscreen keyboard | |||
| Pervasive Sensing to Model Political Opinions in Face-to-Face Networks | | BIBA | Full-Text | 214-231 | |
| Anmol Madan; Katayoun Farrahi; Daniel Gatica-Perez; Alex Pentland | |||
| Exposure and adoption of opinions in social networks are important questions in education, business, and government. We describe a novel application of pervasive computing based on using mobile phone sensors to measure and model the face-to-face interactions and subsequent opinion changes amongst undergraduates, during the 2008 US presidential election campaign. We find that self-reported political discussants have characteristic interaction patterns and can be predicted from sensor data. Mobile features can be used to estimate unique individual exposure to different opinions, and help discover surprising patterns of dynamic homophily related to external political events, such as election debates and election day. To our knowledge, this is the first time such dynamic homophily effects have been measured. Automatically estimated exposure explains individual opinions on election day. Finally, we report statistically significant differences in the daily activities of individuals that change political opinions versus those that do not, by modeling and discovering dominant activities using topic models. We find people who decrease their interest in politics are routinely exposed (face-to-face) to friends with little or no interest in politics. | |||
| Lessons from Touring a Location-Based Experience | | BIBAK | Full-Text | 232-249 | |
| Leif Oppermann; Martin Flintham; Stuart Reeves; Steve Benford; Chris Greenhalgh; Joe Marshall; Matt Adams; Ju Row-Farr; Nick Tandavanitj | |||
| Touring location-based experiences is challenging as both content and
underlying location-services must be adapted to each new setting. A study of a
touring performance called Rider Spoke as it visited three different cities
reveals how professional artists developed a novel approach to these challenges
in which users drove the co-evolution of content and the underlying
location-service as they explored each new city. We show how the artists
iteratively developed filtering, survey, visualization and simulation tools and
processes to enable them to tune the experience to the local characteristics of
each city. Our study reveals how by paying attention to both content and
infrastructure issues in tandem the artists were able to create a powerful user
experience that has since toured to many different cities. Keywords: Location-based performance; cycling; adaptation; Wi-Fi fingerprinting;
seams; user generated content | |||
| Hybrid Prototyping by Using Virtual and Miniature Simulation for Designing Spatial Interactive Information Systems | | BIBAK | Full-Text | 250-257 | |
| Yasuto Nakanishi; Koji Sekiguchi; Takuro Ohmori; Soh Kitahara; Daisuke Akatsuka | |||
| In this paper, we introduce CityCompiler, an integrated environment for the
iteration-based development of spatial interactive systems. CityCompiler
visualizes interactive systems in a virtual 3D space by combining the
Processing source code and the 3D model of the real space, designed with Google
SketchUp. A simulation in virtual space enables us to test a spatial layout and
a combination of components. In addition, the system function of smoothly
switching between a virtual sensor and a real sensor realizes hybrid
prototyping by means of virtual simulation and miniature simulation. These
features integrate the space design with the software design and allow the
smooth deployment of spatial interactive information systems into the real
world. Keywords: software design; space design; prototyping; deployment; IDE | |||
| Designing Shared Public Display Networks -- Implications from Today's Paper-Based Notice Areas | | BIBAK | Full-Text | 258-275 | |
| Florian Alt; Nemanja Memarovic; Ivan Elhart; Dominik Bial; Albrecht Schmidt; Marc Langheinrich; Gunnar Harboe; Elaine M. Huang; Marcello P. Scipioni | |||
| Large public displays have become a regular conceptual element in many shops
and businesses, where they advertise products or highlight upcoming events. In
our work, we are interested in exploring how these isolated display solutions
can be interconnected to form a single large network of public displays, thus
supporting novel forms of sharing access to display real estate. In order to
explore the feasibility of this vision, we investigated today's practices
surrounding shared notice areas, i.e. places where customers and visitors can
put up event posters and classifieds, such as shop windows or notice boards. In
particular, we looked at the content posted to such areas, the means for
sharing it (i.e., forms of content control), and the reason for providing the
shared notice area. Based on two-week long photo logs and a number of in-depth
interviews with providers of such notice areas, we provide a systematic
assessment of factors that inhibit or promote the shared use of public display
space, ultimately leading to a set of concrete design implication for providing
future digital versions of such public notice areas in the form of networked
public displays. Keywords: public display; observation; advertising | |||
| Recognizing the Use of Portable Electrical Devices with Hand-Worn Magnetic Sensors | | BIBAK | Full-Text | 276-293 | |
| Takuya Maekawa; Yasue Kishino; Yasushi Sakurai; Takayuki Suyama | |||
| The new method proposed here recognizes the use of portable electrical
devices such as digital cameras, cellphones, electric shavers, and video game
players with hand-worn magnetic sensors by sensing the magnetic fields emitted
by these devices. Because we live surrounded by large numbers of electrical
devices and frequently use these devices, we can estimate high-level daily
activities by recognizing the use of electrical devices. Therefore, many
studies have attempted to recognize the use of electrical devices with such
approaches as ubiquitous sensing and infrastructure-mediated sensing. A feature
of our method is that we can recognize the use of electrical devices that are
not connected to the home infrastructure without the need for any ubiquitous
sensors attached to the devices. We evaluated the performance of our
recognition method in real home environments, and confirmed that we could
achieve highly accurate recognition with small numbers of hand-worn magnetic
sensors. Keywords: Wearable sensing; Activity recognition; Magnetic sensor | |||
| 3D Gesture Recognition: An Evaluation of User and System Performance | | BIBAK | Full-Text | 294-313 | |
| Michael Wright; Chun-Jung Lin; Eamonn O'Neill; Darren Cosker; Peter Johnson | |||
| We report a series of empirical studies investigating gesture as an
interaction technique in pervasive computing. In our first study, participants
generated gestures for given tasks and from these we identified archetypal
common gestures. Furthermore, we discovered that many of these user-generated
gestures were performed in 3D. We implemented a computer vision based 3D
gesture recognition system and applied it in a further study in which
participants used the common gestures generated in the first study. We
investigated the trade off between system performance and human performance and
preferences, deriving design recommendations. We achieved 84% recognition
accuracy by our prototype 3D gesture recognition system after tuning it through
the use of simple heuristics. The most popular gestures from Study 1 were
regarded by participants in Study 2 as best matching the task they represented,
and they produced the fewest recall errors. Keywords: Gestural interaction; 3D gesture recognition | |||
| Recognition of Hearing Needs from Body and Eye Movements to Improve Hearing Instruments | | BIBAK | Full-Text | 314-331 | |
| Bernd Tessendorf; Andreas Bulling; Daniel Roggen; Thomas Stiefmeier; Manuela Feilner; Peter Derleth; Gerhard Tröster | |||
| Hearing instruments (HIs) have emerged as true pervasive computers as they
continuously adapt the hearing program to the user's context. However, current
HIs are not able to distinguish different hearing needs in the same acoustic
environment. In this work, we explore how information derived from body and eye
movements can be used to improve the recognition of such hearing needs. We
conduct an experiment to provoke an acoustic environment in which different
hearing needs arise: active conversation and working while colleagues are
having a conversation in a noisy office environment. We record body movements
on nine body locations, eye movements using electrooculography (EOG), and sound
using commercial HIs for eleven participants. Using a support vector machine
(SVM) classifier and person-independent training we improve the accuracy of 77%
based on sound to an accuracy of 92% using body movements. With a view to a
future implementation into a HI we then perform a detailed analysis of the
sensors attached to the head. We achieve the best accuracy of 86% using eye
movements compared to 84% for head movements. Our work demonstrates the
potential of additional sensor modalities for future HIs and motivates to
investigate the wider applicability of this approach on further hearing
situations and needs. Keywords: Hearing Instrument; Assistive Technology; Activity Recognition;
Electrooculography (EOG) | |||
| Recognizing Whether Sensors Are on the Same Body | | BIBA | Full-Text | 332-349 | |
| Cory Cornelius; David Kotz | |||
| As personal health sensors become ubiquitous, we also expect them to become
interoperable. That is, instead of closed, end-to-end personal health sensing
systems, we envision standardized sensors wirelessly communicating their data
to a device many people already carry today, the cellphone. In an open personal
health sensing system, users will be able to seamlessly pair off-the-shelf
sensors with their cellphone and expect the system to just work. However, this
ubiquity of sensors creates the potential for users to accidentally wear
sensors that are not necessarily paired with their own cellphone. A husband,
for example, might mistakenly wear a heart-rate sensor that is actually paired
with his wife's cellphone. As long as the heart-rate sensor is within
communication range, the wife's cellphone will be receiving heart-rate data
about her husband, data that is incorrectly entered into her own health record.
We provide a method to probabilistically detect this situation. Because accelerometers are relatively cheap and require little power, we imagine that the cellphone and each sensor will have a companion accelerometer embedded with the sensor itself. We extract standard features from these companion accelerometers, and use a pair-wise statistic -- coherence, a measurement of how well two signals are related in the frequency domain -- to determine how well features correlate for different locations on the body. We then use these feature coherences to train a classifier to recognize whether a pair of sensors -- or a sensor and a cellphone -- are on the same body. We evaluate our method over a dataset of several individuals walking around with sensors in various positions on their body and experimentally show that our method is capable of achieving an accuracies over 80%. | |||
| Sensing and Classifying Impairments of GPS Reception on Mobile Devices | | BIBA | Full-Text | 350-367 | |
| Henrik Blunck; Mikkel Baun Kjærgaard; Thomas Skjødeberg Toftegaard | |||
| Positioning using GPS receivers is a primary sensing modality in many areas of pervasive computing. However, previous work has not considered how people's body impacts the availability and accuracy of GPS positioning and for means to sense such impacts. We present results that the GPS performance degradation on modern smart phones for different hand grip styles and body placements can cause signal strength drops as high as 10-16 dB and double the positioning error. Furthermore, existing phone applications designed to help users identify sources of GPS performance impairment are restricted to show raw signal statistics. To help both users as well as application systems in understanding and mitigating body and environment-induced effects, we propose a method for sensing the current sources of GPS reception impairment in terms of body, urban and indoor conditions. We present results that show that the proposed autonomous method can identify and differentiate such sources, and thus also user environments and phone postures, with reasonable accuracy, while relying solely on GPS receiver data as it is available on most modern smart phones. | |||