User's Current Activity Research Articles

A User-Centric approach for Personalization based on Human Activity Recognition.

The objective of this work focuses on multiple independent user profiles that capture behavioral, emotional, medical, and physical patterns in the working and living environment resulting in one general user profile. Depending on the user's current activity (e.g. walking, eating, etc.), medical history, and other influential factors, the developed framework acts as a supplemental assistant to the user by providing not only the ability to enable supportive functionalities (e.g. image filtering, magnification, etc.) but also informative recommendations (e.g. diet, alcohol, etc.). The personalization of such a profile lies within the user's past preferences using human activity recognition as a base, and it is achieved through a statistical model, the Bayesian belief network. Training and real-time methodological pipelines are introduced and validated. The employed deep learning techniques for identifying human activities are presented and validated in publicly available and in-house datasets. The overall accuracy of human activity recognition reaches up to 86.96 %.

Forward Chaining for Contextual Music Recommendation System

Music is an important aspect of people's daily lives. The reasons people listen to music include to fill their free time and to keep the mood in good condition. Music recommendations are a recommendation system that exists not only because of the many types of music available, but also because people's perceptions of music are still not fully understood. But with so many music choices it makes it difficult for users to find music that fits their context. Examples include considering music based on the current user's location or current activities. A system is required that can recommend music in the context faced by the user.Music Recommendation System Development, Based on user context is a mobile application that uses the Android operating system. The recommendations provided by this system use expert system methods with forward chaining flow. The system will process inputs obtained from users and provide musical recommendations in accordance with the references provided by experts. The result of this study is a rule that is built to produce an average accuracy between user choice and system recommendations of 72%.

Gretchen Is: On Your Facebook Profile, Analyzing the Sociolinguistics of Your Status Message

Some Facebook status messages get dozens of comments and "likes", while others are not interacted with at all. But is this social success just good luck, or are there certain features that can allow us to predict whether a status will be popular or not? This presentation examines a cross‐section of authentic Facebook status messages in an attempt to figure out what makes a status socially successful. Three types of status message are identified: the narcissistic status, the informative status, and the shared‐reference status. The narcissistic status is the stereotypical expression of emotion or mundane activity, where the audience is less important, and consequently less engaged, than the author. The informative status provides basic information about the user's current location or activity, which can allow him or her to be more easily contacted by friends, but does not tend to inspire much conversation. However, it is the shared‐reference status, which refers to an experience that the author shares with a relatively small group of friends, that inspires the greatest levels of dialogue in the form of comments and "likes." Further details to be discussed include the varieties of shared‐reference status and the relation of Facebook statuses to external perception of them and to real‐life rules of social interaction.

Gesteme-free context-aware adaptation of robot behavior in human-robot cooperation.

Cooperative robotics is receiving greater acceptance because the typical advantages provided by manipulators are combined with an intuitive usage. In particular, hands-on robotics may benefit from the adaptation of the assistant behavior with respect to the activity currently performed by the user. A fast and reliable classification of human activities is required, as well as strategies to smoothly modify the control of the manipulator. In this scenario, gesteme-based motion classification is inadequate because it needs the observation of a wide signal percentage and the definition of a rich vocabulary. In this work, a system able to recognize the user's current activity without a vocabulary of gestemes, and to accordingly adapt the manipulator's dynamic behavior is presented. An underlying stochastic model fits variations in the user's guidance forces and the resulting trajectories of the manipulator's end-effector with a set of Gaussian distribution. The high-level switching between these distributions is captured with hidden Markov models. The dynamic of the KUKA light-weight robot, a torque-controlled manipulator, is modified with respect to the classified activity using sigmoidal-shaped functions. The presented system is validated over a pool of 12 näive users in a scenario that addresses surgical targeting tasks on soft tissue. The robot's assistance is adapted in order to obtain a stiff behavior during activities that require critical accuracy constraint, and higher compliance during wide movements. Both the ability to provide the correct classification at each moment (sample accuracy) and the capability of correctly identify the correct sequence of activity (sequence accuracy) were evaluated. The proposed classifier is fast and accurate in all the experiments conducted (80% sample accuracy after the observation of ∼450ms of signal). Moreover, the ability of recognize the correct sequence of activities, without unwanted transitions is guaranteed (sequence accuracy ∼90% when computed far away from user desired transitions). Finally, the proposed activity-based adaptation of the robot's dynamic does not lead to a not smooth behavior (high smoothness, i.e. normalized jerk score <0.01). The provided system is able to dynamic assist the operator during cooperation in the presented scenario.

Subject-independent human activity recognition using Smartphone accelerometer with cloud support

Human activity recognition is an important task in providing contextual user information. In this study, we present a methodology to achieve human activity recognition using a Smartphone accelerometer independent of a subject compared with other user-dependent solutions. The proposed system is composed of four components; a data collector, a data storage cloud, a workstation module and an activity recogniser. The data collector extracts a unique set of defined features from raw data and sends them to the data storage cloud. The workstation module receives the training data from the cloud and generates classification models. The activity recogniser determines the user's current activity based on up-to-date available classifier from the cloud. A prototype is implemented on an android platform to recognise a set of basic daily living activities by placing the Smartphone in different positions to the user and evaluated for offline and online testing to show the scalability and effectiveness.

Crowdsensing in Urban Areas for City-Scale Mass Gathering Management: Geofencing and Activity Recognition

The widespread availability of smartphones today equipped with several physical and virtual sensors allows to directly collect various information about surrounding physical and logical context for different purposes that range from detecting user's current physical activity and also user presence in a designated area, often referred to as geofencing, to determining current social pulse of individuals and entire communities. Mobile crowdsensing seems a promising solution for enabling the design/development and deployment of a wide range of advanced applications in various fields. In particular, public safety, transportation, and energy monitoring and management in urban environments can benefit from mobile crowdsensing in terms of advanced provisioned applications as well as savings of investments in the urban sensing infrastructure. However, enabling those advanced smart urban applications requires complex signal processing, machine learning, and resource management algorithms that are often beyond the skills of many mobile app developers. This paper describes the pivotal relevance of these facilities for mobile crowdsensing applications and presents our open-source solution, called Mobile Sensing Technology (MoST), for activity detection and geofencing, comparing it with the reference implementations provided by Google as part of the Google Play Services library. Experimental results within the testbed framework of a crowd-management application scenario validate MoST design guidelines and demonstrate the general-purpose, unintrusive, and power-efficient characteristics of MoST sensing capabilities.

Key Technology Research Based on the Smart Home BSNs

In recent years, Mobile computing and intelligent household hot topic, more and more get the attention of people. The current smart home systems for self-help to collect information about the indoor environment, at the same time can also interact with the user, but the system when the user needs to send control commands always spend time learning about the current state of household environment, manual control will become a kind of bondage. So this paper using a user activity recognition system based on smart phone, its main function is based on the current user on sensor information to obtain the user's current activity, the current environment control command. Combining the system and smart home system, family equipment according to the user's activities independently adjust to cooperate with users, users are more focused on the main transaction processing. This paper will be focused on energy saving problem, smart home system based on smart phone user activity recognition and the coordination between them.

Interactive Self‐Organizing Windows

AbstractIn this paper, we present the design and implementation of a dynamic window management technique that changes the perception of windows as fixed‐sized rectangles. The primary goal of self‐organizing windows is to automatically display the most relevant information for a user's current activity, which removes the burden of organizing and arranging windows from the user. We analyze the image‐based representation of each window and identify coherent pieces of information. The windows are then automatically moved, scaled and composed in a content‐aware manner to fit the most relevant information into the limited area of the screen. During the design process, we consider findings from previous experiments and show how users can benefit from our system. We also describe how the immense processing power of current graphics processing units can be exploited to build an interactive system that finds an optimal solution within the complex design space of all possible window transformations in real time.

Context-enhanced interaction techniques for more accessible mobile phones

Modern mobile phones enable users to access a wide variety of information and communication services anytime and anywhere. In order for people with disabilities to benefit from these services, mobile devices must provide accessible user interfaces. Currently, users with motor and visual impairments often have difficulty using mobile devices due to issues such as screen readability and small controls. These problems may be exacerbated when attempting to use such devices in crowded, noisy or otherwise distracting environments. We are developing accessible mobile interaction techniques that adapt the mobile device interface to the user by leveraging information about the user's current location, activity, and level of ability.

Enhancing security using mobility-based anomaly detection in cellular mobile networks

Location information is an important feature in users' profiles in cellular mobile networks. In this paper, by exploiting the location history traversed by a mobile user, two domain-independent online anomaly detection schemes are designed, namely the Lempel-Ziv (LZ)-based and Markov-based detection schemes. The authors focus on the identification of a group of especially harmful internal attackers-masqueraders. For both schemes, cell IDs traversed by each mobile user are extracted as the feature value. Specifically, the mobility pattern of each user is characterized by a high-order Markov model. The LZ-based detection scheme from the well-developed data compression techniques is derived. Moreover, the technique of exponentially weighted moving average is used to modify a user's normal profile dynamically. The user profile can characterize the normal behavior of each user accurately and is sensitive to abnormal changes. For the Markov-based detection scheme, a fixed-order Markov model is used to characterize the normal behavior. Based on the constructed probability transition matrix, the probability of the user's current activity is calculated. A threshold policy is then used in both schemes to determine whether a mobile device is potentially compromised or not. Simulation results are presented to show the effectiveness of the proposed schemes. Moreover, our results show that the LZ-based detection scheme performs better than the Markov-based detection scheme, especially for low-speed mobile users.