Majdi Rawashdeh

An Efficient Activity Recognition Framework: Toward Privacy-Sensitive Health Data Sensing

Recent advances in wireless sensor networks for ubiquitous health and activity monitoring systems have triggered the possibility of addressing human needs in smart environments through recognizing human real-time activities. While the nature of streams in such networks requires efficient recognition techniques, it is also subject to suspicious inference-based privacy attacks. In this paper, we propose a framework that efficiently recognizes human activities in smart homes based on spatiotemporal mining technique. In addition, we propose a technique to enhance the privacy of the collected human sensed activities using a modified version of micro-aggregation approach. An extensive validation of our framework has been performed on benchmark data sets yielding quite promising results in terms of accuracy and privacy-utility tradeoff.

Context-aware multimodal recommendations of multimedia data in cyber situational awareness

The current proliferation of large amounts of multimedia data creates an unprecedented challenge for security analysts in the context of Cyber Situational Awareness. Due to this phenomenal growth of multimedia data, security analysts have to invest enormous time and efforts in filtering and correlating multimedia data in order to make informed decisions about identifying and mitigating threats and vulnerabilities. In particular, analysts have to analyze and interpret diverse multimedia network data with varying contexts in order to find the true evidence of cyber attacks. Considering the multimedia nature of cyber security data, we propose a cloud-assisted recommendation system that can identify and retrieve multimedia data of interest based on contextual information and security analysts’ personal preferences. This recommendation system benefits security analysts by establishing a bridge between their personal preferences, the contextual information of their analytical process, and the various types of modality of multimedia data. Evaluation of the proposed system shows evidence that our multimedia recommendation mechanisms promotes cyber threat understanding and risk assessment.

Mining tag-clouds to improve social media recommendation

Massive amounts of data are available on social websites, therefore finding the suitable item is a challenging issue. According to recent social statistics, we have more than 930 million people are using WhatsApp with more than 340 million active daily users and 955 million people who access Facebook daily with an average daily photo uploads up to 325 million. The approach presented in this paper employs the collaborative tagging accumulated by huge number of users to improve social media recommendation. Our approach has two phases, in the first phase, we compute the tag-item weight model and in the second phase, we compute the user-tag preference model. After that we employ the two models to find the suitable items tailored to the user’s preferences and recommend the items with the highest score. Also our model can compute the tag score and suggest the tags with the highest weight to the user according to their preferences. The experiment results performed on Flicker and MovieLens prove that our approach is capable to improve the social media recommendation.

An ODT-based abstraction for mining closed sequential temporal patterns in IoT-cloud smart homes

Due to the large amount of usage data collected from smart home appliances in an IoT-cloud environment, efficient mining techniques are of great need to capture the behavioral patterns. Existing mining algorithms are time-consuming and error prone as the amount of data is increasing rapidly. In this paper, we propose an abstraction approach to model temporal data based on an ordered decision tree (ODT) and spatiotemporal characteristics of usage data for IoT-cloud paradigm. The contribution of this research is to provide an efficient representation in terms of average length of patterns, while preserving the spatiotemporal characteristics of original data. We performed extensive experiments on synthetic data to report the performance and provide a comparison with state-of-the-art algorithms to prove the correctness of the proposed technique, even at a low-level of abstraction. The results indicate that the proposed methodology outperform existing techniques due to the inherited power of the ODT temporal structure.

Multimedia Mobile Cloud Computing: Application models for performance enhancement

Central to the vision of Smart City is the realization of efficient models that are capable of handling massive amounts of mobile multimedia data in the citys eco-system. Despite the improvement in hardware of mobile devices, challenges associated with analyzing, managing, and sharing of data still exist. As such, the responsiveness to real time applications and bandwidth and wireless constraints cannot be achieved by hardware design only. Therefore, there is a move towards the software side that is enabled by Mobile Cloud Computing to overcome these challenges, whereby parts of mobile applications are executed in remote servers with rich computational resources. This technology decreases the load on mobile devices and increases their performance. Several models have been proposed to increase mobile devices performance. This paper explores these models and compares them based on several performance parameters, including computation offloading, latency, bandwidth, and response time.

Reusing Software Libraries Using Semantic Graphs

This paper describes a two-part system that helps developers understand how to search and reuse complex software libraries. First, we present an approach for automatic retrieval of software components in reuse libraries. Second, the system implements a source-code recommendation approach which automatically constructs and delivers relevant code examples that demonstrate how the retrieved components can be used to solve particular programming problems. In arriving at such solutions, we utilize ontological modeling to provide semantic representation of the conceptual source-code knowledge in software libraries. This representation is the basis for computing entailments and enabling semantic reasoning. Our experiments show evidence that combining ontology formalisms with context-sensitive techniques enhance precision when retrieving and recommending reusable code even without mining a carefully crafted corpus of similar code.

Reliable service delivery in Tele-health care systems

Abstract Modern ICT Applications on Tele-health focuses on providing the smart infrastructure that facilitates the delivery of health services. While Internet-of-Things (IoT) and cloud-computing platforms assist the implementation of such architecture, the reliability of service delivery during network disconnection is still an open issue in this domain. This paper proposes a prediction methodology that is able to deliver reliable services with acceptable accuracy by incorporating domain-specific knowledge into exchanged data. The proposed service will be of a great value in a situation where the network availability is not reliable. The contributions of this work are to 1) measure the impact of ontology enrichment on classifying the health data, 2) develop a prediction model that is able to predict patients readings with an acceptable accuracy, and 3) minimize communicating messages among the network components. Three experiments have been conducted on a real health dataset to measure the performance of the proposed methodology. The results showed that our proposed methodology improved the reliability of the Tele-health services implemented on the top of IoT and cloud-computing platforms.

An Annotation Technique for In-Home Smart Monitoring Environments

Advances in multimedia technologies have led to the emergence of smart home applications. In fact, mobile multimedia technologies provide the infrastructure to adopt smart solutions and track inhabitants’ activities. In-home activity recognition significantly enhances the performance of healthcare-monitoring and emergency-control applications for elderly and people with special needs. Developing and validating data models for such applications requires training sets that reflect a ground truth in the form of labeled or annotated data. With the accelerated development of Internet-of-Things applications, automated annotation processes have emerged understanding resident behavior in terms of activities. This paper presents a methodology for automatic data annotation by profiling sensing nodes. Our proposed methodology models activities based on spatially recognized actions, with every activity expected to have a direct relationship with a specific set of locations. Furthermore, the proposed technique validates the assignment of labels based on the temporal relations among consecutive actions. We performed experiments to evaluate our proposed methodology on CASAS data sets, which indicated that the proposed methodology achieved better performance, to a statistically significant extent, than the state-of-the-art methodologies presented in the literature.

A multimedia cloud-based framework for constant monitoring on obese patients

Obesity phenomenon has become a significant issue over the world. Obesity has various negative consequences that might impact not only the health but also the social and the economic issues. Current studies reveal the lack of patients commitment to the doctors instructions. In this paper, we propose a new cloud-based model with ultimate aim to monitor obese patients health condition and behavior constantly under a real-time vision of the caregiver. The proposed model provides a technical method to record, disseminates, and share knowledge and awareness among patients and caregivers. This model utilizes wireless body sensor network devices to measure heart rate; mobile web-service as a middleware to process the data from/to cloud knowledge-base and a caregiver backend/dashboard for the real-time monitor. A live test demo of the model experimented on 55 subjects to check its applicability and cost-effectiveness. The results were promised and prove its ability to help overcoming this disease.

Models for Multimedia Mobile Cloud in Smart Cities

Realizing efficient models that are capable of handling massive amounts of mobile multimedia data is an essential component of achieving the vision of smart future cities. In order for a model to be truly effective, it can not be based solely on improving the hardware of mobile devices. Instead, utilizing software solutions and tools provided by the emerging field of Mobile Cloud Computing (MCC) provides results that could not be otherwise possible. In MCC, parts of mobile applications are executed in remote servers that provide powerful computational resources. As such, the performance of mobile devices is significantly increased since the heavy computational load is being handled in the cloud. In this paper, we provide a comprehensive analysis and comparisons of current MCC models that are meant to increase mobile devices performance and utility. In particular, we provide a critical discussion related to the applicability of such models, and we compare these models using a devised set of criteria that affect the performance and reliability of effective Mobile Cloud Application models.