Sign In

Communications of the ACM

81 - 90 of 286 for bentley

Sharing (and Discussing) the Moment: The Conversations that Occur Around Shared Mobile Media

Today's smartphones enable rich, media-enhanced conversations. Millions of photos and billions of messages are shared each day on smartphones. But how, exactly, are images and web links being used in mobile conversations? And what does this mean for the design of new mobile communications applications? We set out to learn how people currently share and discuss mobile media by performing a detailed content analysis of 109 photos and links that were shared in 2,779 messages using a mobile messaging application deployed in the United State and Taiwan. Through our analysis of these conversations, we show how mobile media is used to experience the moment together, to fill in the visual details, to provide background context, and to exchange information. We then discuss our results and provide two designs inspired by our findings.


"I'm just on my phone and they're watching TV": Quantifying mobile device use while watching television

In recent years, mobile devices have become a part of our daily lives--much like television sets had over the second half of the 20th century. Increasingly, people are using mobile devices while watching television. We set out to understand this behavior on a minute-by-minute quantified level as well as users' motivations and purposes of device use while watching television. We conducted a novel mixed-methods study inside seven households with fourteen instrumented phone and tablet devices, capturing all app launches and app use durations, correlated with the moment in the television program when they occurred. Surprisingly, we found little difference between the volume of device use during programs and commercials. Our two main findings are that 1) participants often joined family members in the TV room to physically be together; when they lack interest in the program, they spend the majority of the show on a secondary device and watch TV only during key moments. 2) Virtually none of participants' app and web use during TV consumption was directly related to the running show. With our study, we set the stage for larger-scale investigations into the details of mobile interactions while watching television. Our novel method will aid future work of the community as a means of fully understanding multi-device use alongside television consumption.


Reducing the Stress of Coordination: Sharing Travel Time Information Between Contacts on Mobile Phones

We explore the everyday use of a new abstraction for mo-bile location-sharing. By sharing the travel time between contacts calculated for walking, transit, and driving, we have enabled users to more easily coordinate meeting up and planning family obligations. Specifically, our participants reported that the information helped to lower the stress of these activities and provided reassurance of the arrival times of their close friends and family. In this paper, we describe our system, motivate its design, and explore results from a 20-user, 21-day field trial showing the use-fulness of the abstraction as well as attitudes towards privacy when sharing travel times with close friends or family.


The Composition and Use of Modern Mobile Phonebooks

Over the past decade, the mobile phonebook has evolved from a relatively short list of people that one calls and texts to a many-hundred person list of aggregated contacts from around the web. This is happening at a time when an increasing number of mobile applications are relying on the mobile phonebook to create one's social network in their services. Through a large-scale study of the phonebooks of 200 diverse participants, containing 65,940 contacts, we set out to understand today's mobile contact lists. Our participants reported that they did not recognize the names of 29% of their contacts and we found that the most frequently contacted five contacts represent greater than 80% of all calls and text messages with phonebook contacts. We conclude with implications for the design of mobile applications that rely on phonebook data.


"It's kind of like an extra screen for my phone": Understanding Everyday Uses of Consumer Smart Watches

The CHI, Ubicomp, and UIST communities have been studying watch-based interactions for many years. While much of this work has been technical or focused on interaction techniques in the lab, now smart watch devices are available directly to consumers from a variety of manufacturers. However, little has been studied as to why people adopt these devices and the real-world problems that they are solving in their lives. We set out to explore current smart watch use in an interview-based study of five diverse participants. We will use data from this study to help design and develop new smart watch applications.


Dynamic learning of heart sounds with changing noise: an ais-based multi-agent model using systemic computation

Agent-Based Models are used to model dynamic systems such as stock markets, societies, and complex biological systems that are difficult to model analytically using partial differential equations. Many agent-based modeling software are designed for serial von-Neumann computer architectures. That limits the speed and scalability of these systems. Systemic computation (SC) is designed to be a model of natural behavior and, at the same time, a non Von-Neumann architecture with its characteristics similar to multi-agent system. Here we propose a novel method based on an Artificial Immune System (AIS) and implemented on a systemic computer, which is designed to adapt itself over continuous arrival of data to cope with changing patterns of noise without requirement for feedback, as a result of its own experience. Experiments with heartbeat data collected from a clinical trial in hospitals using a digital stethoscope shows the algorithm performs up to 3.60% better in the precision rate of murmur and 3.96% better in the recall rate of murmur than other standard anomaly detector approaches such as Multiple Kernel Anomaly Detection (MKAD).


Adapting to dynamically changing noise during learning of heart sounds: an AIS-based approach using systemic computation

Real world machine learning, where data is sampled continuously, may in theory be classifiable into distinct and unchanging categories but in practice the classification becomes non-trivial because the nature of the background noise continuously changes. Applying distinct and unchanging categories for data ignores the fact that for some applications where the categories of data may remain constant, the background noise constantly changes, and thus the ability for a supervised learning method to work is limited. In this work, we propose a novel method based on an Artificial Immune System (AIS) and implemented on a systemic computer, which is designed to adapt itself over continuous arrival of data to cope with changing patterns of noise without requirement for feedback, as a result of its own experience.


MyChannel: exploring city-based multimedia news presentations on the living room TV

We see the television as a primary device to connect view-ers with the information and people that matter most in their lives. Televisions, as central places where the family gath-ers, provide a unique location to elevate news and social updates that can connect family and friends across a dis-tance. Through creating the MyChannel service, a TV-based personalized news program, we have explored the types of content that work best in this format. We have also gained a detailed understanding of how television content can inspire feelings of connection and communication with friends and family at a distance through an 8-day in-home field evaluation. We describe the system and findings from our studies and close with a discussion on the future of per-sonalized television news.


Load balancing n-body simulations with highly non-uniform density

N-body methods simulate the evolution of systems of particles (or bodies). They are critical for scientific research in fields as diverse as molecular dynamics, astrophysics, and material science. Most load balancing techniques for N-body methods use particle count to approximate computational work. This approximation is inaccurate, especially for systems with high density variation, because work in an N-body simulation is proportional to the particle density, not the particle count. In this paper, we demonstrate that existing techniques do not perform well at scale when particle density is highly non-uniform, and we propose a load balance technique that efficiently assigns load in terms of interactions instead of particles. We use adaptive sampling to create an even work distribution more amenable to partitioning, and to reduce partitioning overhead. We implement and evaluate our approach on a Barnes-Hut algorithm and a large-scale dislocation dynamics application, ParaDiS. Our method achieves up to 26% improvement in overall performance of Barnes-Hut and 18% in ParaDiS.