Generating Natural Language Explanations (NLEs) for model predictions on medical images, particularly those depicting thoracic pathologies, remains a critical and challenging task. Existing methodologies often struggle due to general models' insufficient domain-specific medical knowledge and privacy concerns associated with retrieval-based augmentation techniques. To address these issues, we propose a novel Vision-Language framework augmented with a Knowledge Graph (KG)-based datastore, which enhances the model's understanding by incorporating additional domain-specific medical knowledge essential for generating accurate and informative NLEs. Our framework employs a KG-based retrieval mechanism that not only improves the precision of the generated explanations but also preserves data privacy by avoiding direct data retrieval. The KG datastore is designed as a plug-and-play module, allowing for seamless integration with various model architectures. We introduce and evaluate three distinct frameworks within this paradigm: KG-LLaVA, which integrates the pre-trained LLaVA model with KG-RAG; Med-XPT, a custom framework combining MedCLIP, a transformer-based projector, and GPT-2; and Bio-LLaVA, which adapts LLaVA by incorporating the Bio-ViT-L vision model. These frameworks are validated on the MIMIC-NLE dataset, where they achieve state-of-the-art results, underscoring the effectiveness of KG augmentation in generating high-quality NLEs for thoracic pathologies.

Large language models (LLMs) have significantly improved numerous natural language processing tasks. However, their performance relies heavily on the provided instructions or prompts. Recently, several prompting methodologies have been developed to enhance the reasoning abilities of LLMs. Notably, the Chain-of-Thought (CoT) approach provides examples that help break down tasks into sub-steps, resulting in more accurate solutions. However, the process of generating detailed examples may not be user-friendly, as end users prefer providing task descriptions rather than a set of examples. In this study, we introduce Chain-of-Factors (CoF), an innovative zero-shot prompting methodology that incorporates task-specific instructions as a chain of factors into the prompt, aimed at enhancing the factor-centric reasoning abilities of LLMs. Experiments on three LLMs, including ChatGPT-3.5, Gemini, and GPT-4, show performance improvements ranging from 0.01% to 40.2% in accuracy on various symbolic reasoning and logical reasoning tasks compared with zero-shot and few-shot CoT. In summary, CoF enhances LLMs' reasoning abilities by including task-specific steps and instructions, while also decreasing the necessity for fine-tuning specific to each task.

Abstract Edge computing, a distributed computing architecture within the knowledge-defined network (KDN), faces challenges due to the significant disparities and data heterogeneity among its nodes, hindering their interaction. Ontology, a solution within the Semantic Web, is well-suited for addressing data heterogeneity and matching ontologies effectively. However, ontology matching presents difficulties due to non-linear mathematical issues. To overcome these challenges, the generative adversarial network (GAN), an unsupervised learning method, has emerged as a promising tool. GAN consists of two models with distinct objectives trained against eachother to achieve optimal outcomes. This paper introduces SA-GAN, an algorithm that combines GAN with simulation-based annealing to enhance its effectiveness. SA-GAN utilizes a stagnation counter to expedite the convergence speed of GAN. Through experiments conducted on a renowned ontology benchmark, the paper demonstrates that SA-GAN, along with other ontology matching algorithms, can identify the best alignments. Consequently, SA-GAN facilitates the construction of bridges in edge computing, improving its overall effectiveness.

Abstract Chronic Diseases are a prevalent problem that affects millions of people worldwide. It is a prevalent health condition that requires careful diet and medication management and preventing chronic diseases. Traditional approaches to nutritional recommendation generation often rely on generic guidelines and population-based data, which may not account for individual dietary needs and preferences variations. In this paper, we propose a knowledge graph driven methodology for generating highly personalized nutritional recommendations that leverage the power of knowledge graphs to integrate and analyze complex data about an individual's health, lifestyle, and dietary habits. Our methodology employs a multi-step process that includes data collection and curation, knowledge graph construction, and personalized recommendation generation. We illustrate the effectiveness of our approach through a case study in which we generate personalized nutritional recommendations for a sample individual based on their specific health and dietary goals.

OBJECTIVE: With the scenario of limited labeled dataset, this paper introduces a deep learning-based approach that leverages Diabetic Retinopathy (DR) severity recognition performance using fundus images combined with wide-field swept-source optical coherence tomography angiography (SS-OCTA). METHODS: The proposed architecture comprises a backbone convolutional network associated with a Twofold Feature Augmentation mechanism, namely TFA-Net. The former includes multiple convolution blocks extracting representational features at various scales. The latter is constructed in a two-stage manner, i.e., the utilization of weight-sharing convolution kernels and the deployment of a Reverse Cross-Attention (RCA) stream. RESULTS: The proposed model achieves a Quadratic Weighted Kappa rate of 90.2% on the small-sized internal KHUMC dataset. The robustness of the RCA stream is also evaluated by the single-modal Messidor dataset, of which the obtained mean Accuracy (94.8%) and Area Under Receiver Operating Characteristic (99.4%) outperform those of the state-of-the-arts significantly. CONCLUSION: Utilizing a network strongly regularized at feature space to learn the amalgamation of different modalities is of proven effectiveness. Thanks to the widespread availability of multi-modal retinal imaging for each diabetes patient nowadays, such approach can reduce the heavy reliance on large quantity of labeled visual data. SIGNIFICANCE: Our TFA-Net is able to coordinate hybrid information of fundus photos and wide-field SS-OCTA for exhaustively exploiting DR-oriented biomarkers. Moreover, the embedded feature-wise augmentation scheme can enrich generalization ability efficiently despite learning from small-scale labeled data.

Abstract The case‐based learning (CBL) approach has gained attention in medical education as an alternative to traditional learning methodology. However, current CBL systems do not facilitate and provide computer‐based domain knowledge to medical students for solving real‐world clinical cases during CBL practice. To automate CBL, clinical documents are beneficial for constructing domain knowledge. In the literature, most systems and methodologies require a knowledge engineer to construct machine‐readable knowledge. Keeping in view these facts, we present a knowledge construction methodology (KCM‐CD) to construct domain knowledge ontology (i.e., structured declarative knowledge) from unstructured text in a systematic way using artificial intelligence techniques, with minimum intervention from a knowledge engineer. To utilize the strength of humans and computers, and to realize the KCM‐CD methodology, an interactive case‐based learning system(iCBLS) was developed. Finally, the developed ontological model was evaluated to evaluate the quality of domain knowledge in terms of coherence measure. The results showed that the overall domain model has positive coherence values, indicating that all words in each branch of the domain ontology are correlated with each other and the quality of the developed model is acceptable.

Over the past few decade, smartphone technology has played a vital role in the healthcare domain. In this paper, we proposed a methodology to allow end users to automatically process the medical text image using a camera or enter text manually. Then the system output allows to extract medical concepts, its semantic type and entity type from medical text apply Natural Processing Language techniques from UMLS medical dictionary. The medical text can be a health report, a clinical case, or other kinds of a text containing medically related information. The aim of this kind of methodology of the mobile application is to contribute in the area of natural language processing (NLP), intelligent system and to increase the medical students and bioinformatics researchers interest to quickly accessing information about medical data.

Wellness platform plays a vital role to prevent chronic disease. In this paper, we have introduced wellness smartphone application within the ambit of the Mining Mind project, which aims to support the people to adopt healthy behavior and lifestyle. When an unhealthy behavior is detected, personalized physical recommendations are generated automatically by the Mining Mind wellness platform. Recommendations are delivered through Push notification and display on the application main screen for the user. The application has feedback functionalities on the effectiveness of recommendation and education. The application also supports descriptive analytics of wellness goals achieved on a monthly, weekly and daily basis.

Smart gadgets play a vital role to adapt the unhealthy lifestyle by converging science and technology in this digital world. The challenge of behavior change through step-by-step coaching and guidance is realized by just-in-time intervention using mobile computing. The amalgamation of behavior change theories with digital technologies support us to mold the behavior in a scientific manner. Wellness platform based behavior analysis is performed through unbiased life-log as well as questionnaire for permanent and ad-hoc users. Daily personalized coaching has been provided to motivate the users, whereas instant context-based recommendations have supported stage-wise adaptation of healthy behavior.

We have designed and implement a random text dependent voice based authentication protocol for smartphones. The objective was to provide an efficient and reliable authentication mechanism that ensure prevention against the emerging attacks. In this paper, we have focused on the architecture, protocol, and prevention against replay attack only.

OBJECTIVE: The study sought to develop a clinical decision support system (CDSS) for the treatment of thyroid nodules, using a mind map and iterative decision tree (IDT) approach to the integration of clinical practice guidelines (CPGs). MATERIALS AND METHODS: Thyroid nodule CPGs of the American Thyroid Association and Korean Thyroid Association were analyzed by endocrine surgeons (domain experts) and computer scientists. Clinical knowledge from the CPGs was expressed using mind maps. The mind maps were analyzed and converted into IDTs. The final IDT was implemented as a set of candidate rules (3700) for a knowledge-based CDSS. The system was evaluated via a retrospective review of the medical records of 483 patients who had undergone thyroidectomy between January and December 2015 at a single tertiary center (Seoul National University Hospital Bundang, Korea). RESULTS: Concordance between CDSS recommendations and treatment in routine clinical practice was 78.9%. In the 21.1% discordant cases, deviation from the CDSS treatment recommendation was mainly attributable to (1) refusal of the patient to undergo total thyroidectomy and (2) conversion from lobectomy to total thyroidectomy following an unexpected histological finding during intraoperative frozen biopsy lymph node analysis. CONCLUSIONS: The present study demonstrated that a knowledge-based CDSS is feasible in the treatment of thyroid nodules. A high-quality knowledge-based CDSS was developed, and medical domain and computer scientists collaborated effectively in an integrated development environment. The mind map and IDT approach represents a pioneering method of integrating knowledge from CPGs.

This paper addresses the problem of querying Knowledge bases (KBs) that store semantic big data. For efficiently querying data the most important factor is cache replacement policy, which determines the overall query response. As cache is limited in size, less frequently accessed data should be removed to provide more space to hot triples (frequently accessed). So, to achieve a similar performance to RDBMS, we proposed an Adaptive Cache Replacement (ACR) policy that predict the hot triples from query log. Moreover, performance bottleneck of triplestore, makes realworld application difficult. To achieve a closer performance similar to RDBMS, we have proposed an Adaptive Cache Replacement (ACR) policy that predict the hot triples from query log. Our proposed algorithm effectively replaces cache with high accuracy. To implement cache replacement policy, we have applied exponential smoothing, a forecast method, to collect most frequently accessed triples. The evaluation result shows that the proposed scheme outperforms the existing cache replacement policies, such as LRU (least recently used) and LFU (least frequently used), in terms of higher hit rates and less time overhead.

Clinical Practice Guidelines (CPGs) are considered as an effective tool to improve and standardize healthcare. A number of organizations are developing and maintaining clinical guidelines to provide state of the art healthcare services. However, the guidelines contain background information along with disease specific information which tends to create difficulties in using it during actual practice as well as in transforming it into a machine interpretable format. The most relevant information needs to be isolated from irrelevant information. In this study, we proposed a methodology that separates relevant information known as recommendation statements from irrelevant information by using heuristic patterns. We have extracted 10 patterns in a semi-automatic manner from hypertension guideline and evaluated the extracted patterns for identifying recommendation statements in the guideline and achieved 85.54% accuracy. These extracted patterns facilitate domain expert to get disease specific information in real time during the clinical workflow. Moreover, it can also work as a preprocessing step during the transformation of guideline to computer interpretable models.

Clinical decision systems (CDSs) showed promising results in different areas including reminder systems, diagnostic, drug dose, prescription, and other pharma related domains. However, many aspects are subject to rethink and redesign in the era of precision medicine. Current clinical decision systems are more focusing to serve one or the other area of clinical care, imparting individualistic characteristics of a single and probably an isolated domain or sub-domain. While precision medicine requires a comprehensive data and knowledge for making precise decisions. The comprehensive knowledge shall contain information about disease sub-types, disease risk, diagnosis, therapy, and prognosis. Building such a comprehensive knowledge and executing queries to get the results from the decision support system may require federation at query level and integration at data level that is accumulated from different databases situated in one or more than one setup. In this paper, we research to provide an initial idea and guide of redesigning the CDS architecture in a way to serve the very need of precision medicine. We describe the limitation of existing CDSs, challenges to address them and propose a solution to address those challenges. This work may lay down a foundation for the architectures of futuristic CDSs in the era of precision medicine.

Abstract A physical activity recommendation system promotes active lifestyles for users. Real‐world reasoning and recommendation systems face the issues of data and knowledge integration, knowledge acquisition, and accurate recommendation generation. The knowledge‐based reasoning and recommendation framework (KRF) proposed here, which accurately generates reliable recommendations and educational facts for users, could solve those issues. The KRF methodology focuses on integrating data with knowledge, rule‐based reasoning, and conflict resolution. The integration issue is resolved using a semi‐automatic mapping approach in which rule conditions are mapped to data schema. The rule‐based reasoning methodology uses explicit rules with a maximum‐specificity conflict resolution strategy to ensure the generation of appropriate and correct recommendations. The data used during the reasoning process are generated in real time from users' physical activities and personal profiles in order to personalize recommendations. The proposed KRF is part of a wellness and health care platform, Mining Minds, and has been tested in the Mining Minds integrated environment using a sedentary user behaviour scenario. To evaluate the KRF methodology, a stand‐alone, open‐source application (Version 1.0) was released and tested using a dataset of 10 volunteers with 40 different types of sedentary behaviours. The KRF performance was measured using average execution time and recommendation accuracy.

In this paper, we have evaluated the effectiveness of scheduling strategies in Linked Open Data based application for keeping local data caches up-to-date. We argue that the healthcare organizations are publishing data publicly, but consuming of data is difficult due to rapid growth of the linked data cloud. Most of the applications that are consuming linked data suffer from challenges such as change estimation and accuracy of index for keeping the data fresh for visualization. In this work, we have evaluated the quality of the data updates performed by the scheduling strategies. We have implemented the state-of-the-art web scheduling approaches; ChangeRatio and ChangeRate on linked dataset. We have concluded our evaluation that the strategies based on ChangeRate performed better than the ChangeRatio.

The combination of ontology based context-awareness and machine learning context classification is an interesting research area. The determined contexts are obtained using semantic reasoning based on context ontology developed by expert using domain specific rules. This reasoning suffer challenges of soundness and completeness in real-time deployment. This paper addresses the aforementioned challenges from semantic reasoning by embracing machine learning modeling and classification benefits. Machine learning relies on data, for this we developed training and deployment phase for ontological ABox assertions. Approximately 99.99% precision through machine learning approach was achieved over 91.5% accuracy with semantic reasoning. The statistical evaluation proves the improvement in terms of accuracy for context prediction and overall performance.

This paper presents a robust foreground detection method capable of adapting to different motion speeds in scenes. A key contribution of this paper is the background estimation using a proposed novel algorithm, neighbor-based intensity correction (NIC), that identifies and modifies the motion pixels from the difference of the background and the current frame. Concretely, the first frame is considered as an initial background that is updated with the pixel intensity from each new frame based on the examination of neighborhood pixels. These pixels are formed into windows generated from the background and the current frame to identify whether a pixel belongs to the background or the current frame. The intensity modification procedure is based on the comparison of the standard deviation values calculated from two pixel windows. The robustness of the current background is further measured using pixel steadiness as an additional condition for the updating process. Finally, the foreground is detected by the background subtraction scheme with an optimal threshold calculated by the Otsu method. This method is benchmarked on several well-known data sets in the object detection and tracking domain, such as CAVIAR 2004, AVSS 2007, PETS 2009, PETS 2014, and CDNET 2014. We also compare the accuracy of the proposed method with other state-of-the-art methods via standard quantitative metrics under different parameter configurations. In the experiments, NIC approach outperforms several advanced methods on depressing the detected foreground confusions due to light artifact, illumination change, and camera jitter in dynamic scenes.

Sparseness of user-to-item rating data is one of the major factors that deteriorate the quality of recommender system. To handle the sparsity problem, several recommendation techniques have been proposed that additionally consider auxiliary information to improve rating prediction accuracy. In particular, when rating data is sparse, document modeling-based approaches have improved the accuracy by additionally utilizing textual data such as reviews, abstracts, or synopses. However, due to the inherent limitation of the bag-of-words model, they have difficulties in effectively utilizing contextual information of the documents, which leads to shallow understanding of the documents. This paper proposes a novel context-aware recommendation model, convolutional matrix factorization (ConvMF) that integrates convolutional neural network (CNN) into probabilistic matrix factorization (PMF). Consequently, ConvMF captures contextual information of documents and further enhances the rating prediction accuracy. Our extensive evaluations on three real-world datasets show that ConvMF significantly outperforms the state-of-the-art recommendation models even when the rating data is extremely sparse. We also demonstrate that ConvMF successfully captures subtle contextual difference of a word in a document. Our implementation and datasets are available at http://dm.postech.ac.kr/ConvMF.

Understanding video-based activities have remained the challenge regardless of efforts from the image processing and artificial intelligence community. However, the rapid developing of computer vision in 3D area has brought an opportunity for the human pose estimation and so far for the activity recognition. In this research, the authors suggest an impressive approach for understanding daily life activities in the indoor using the skeleton information collected from the Microsoft Kinect device. The approach comprises two significant components as the contribution: the pose-based feature extraction under the spatio-temporal relation and the topic model based learning. For extracting feature, the distance between two articulated points and the angle between horizontal axis and joint vector are measured and normalized on each detected body. A codebook is then constructed using the K-means algorithm to encode the merged set of distance and angle. For modeling activities from sparse features, a hierarchical model developed on the Pachinko Allocation Model is proposed to describe the flexible relationship between features - poselets - activities in the temporal dimension. Finally, the activities are classified by using three different state-of-the-art machine learning techniques: Support Vector Machine, K-Nearest Neighbor, and Random Forest. In the experiment, the proposed approach is benchmarked and compared with existing methods in the overall classification accuracy.

The number of applications based on Apache Hadoop is dramatically increasing due to the robustness and dynamic features of this system. At the heart of Apache Hadoop, the Hadoop Distributed File System (HDFS) provides the reliability and high availability for computation by applying a static replication by default. However, because of the characteristics of parallel operations on the application layer, the access rate for each data file in HDFS is completely different. Consequently, maintaining the same replication mechanism for every data file leads to detrimental effects on the performance. By rigorously considering the drawbacks of the HDFS replication, this paper proposes an approach to dynamically replicate the data file based on the predictive analysis. With the help of probability theory, the utilization of each data file can be predicted to create a corresponding replication strategy. Eventually, the popular files can be subsequently replicated according to their own access potentials. For the remaining low potential files, an erasure code is applied to maintain the reliability. Hence, our approach simultaneously improves the availability while keeping the reliability in comparison to the default scheme. Furthermore, the complexity reduction is applied to enhance the effectiveness of the prediction when dealing with Big Data.

In recent years, the data growth rate has been observed growing at a staggering rate. Considering data search as a primitive operation and to optimize this process on large volume of data, various solution have been evolved over a period of time. Other than finding the precise similarity, these algorithms aim to find the approximate similarities and arrange them into bins. Locality sensitive hashing (LSH) is one such algorithm that discovers probable similarities prior calculating the exact similarity thus enhance the overall search process in high dimensional search space. Realizing same strategy for encrypted data and that too in public cloud introduces few challenges to be resolved before probable similarity discovery. To address these issues and to formalize a similar strategy like LSH, in this paper we have formalized a technique O-Bin that is designed to work over encrypted data in cloud. By exploiting existing cryptographic primitives, O-Bin preserves the data privacy during the similarity discovery for the binning process. Our experimental evaluation for O-Bin produces results similar to LSH for encrypted data.

Personalized healthcare envisions providing customized treatment and management plans to individuals at their doorstep. Key factors to ensure personalized healthcare is to involve with the individual in their daily life activities and process the gathered information to provide recommendations. We identified the mostly exposed domains for gathering chronic disease patients information that includes: clinical, social media, and daily life activities. Clinical data is related to the health-care of the patients while social media, sensory, and wearables data is related to the wellness data of the patients. A framework is required to monitor the health and wellness information of the patients for health and wellness analytics provisioning to the physicians for better decision making. We propose Personalized, Ubiquitous Life-care Decision Support System (PULSE); a state of the art decision support system that helps physicians and patients in life-style management of chronic disease patients such as Diabetes. The proposed approach not only utilizes clinical information but also personalized information by correlation to find hidden information using big data health analytic for improvement of life-care. PULSE provides health analytics by utilizing and processing clinical information of the patient. In the same way, it provides wellness analytics to the patients by using their social, activities, emotions and daily life information. The co-relation between clinical and personalized analytics is performed for better recommendations to the patients. This eventually results in improved life-care and healthy living of the individuals.

A trust-aware recommender system (TARS) is widely used in social media to find useful information. S_Searching is one of the most effective recommender searching mechanisms of TARS. It is based on the scale-freeness of the trust network: a skeleton, which consists of hub nodes of the trust network, is involved in trust propagations. Trusts are first propagated from active users to the skeleton, and then recommenders are searched via the skeleton. One fundamental research issue in S_Searching is to search the skeleton for active users efficiently. Existing methods fully search the trust network to find hub nodes in the skeleton for active users. It has high computational cost. In this paper, we propose a novel iterative deepening-based skeleton searching strategy for S_Searching, in which a depth-limited search is run repeatedly. The depth limit is increased with each iteration until it reaches the maximum allowable trust propagation distance. Simulation results show that the computational complexity of our proposed strategy is much less expensive than that of existing methods.

Summary It is a fact that the attention of research community in computer science, business executives, and decision makers is drastically drawn by big data. As the volume of data becomes bigger, it needs performance‐oriented data‐intensive processing frameworks such as MapReduce, which can scale computation on large commodity clusters. Hadoop MapReduce processes data in Hadoop Distributed File System as jobs scheduled according to YARN fair scheduler and capacity scheduler. However, with advancement and dynamic changes in hardware and operating environments, the performance of clusters is greatly affected. Various efforts in literature have been made to address the issues of heterogeneity (i.e., clusters consisting of virtual machines and machines with different hardware), network communication, data locality, better resource utilization, and run‐time scheduling. In this paper, we present a survey to discuss various research efforts made so far to improve Hadoop MapReduce scheduling. We classify scheduling algorithms and techniques proposed in the literature so far based on their addressing areas and present a taxonomy. Furthermore, we also discuss various aspects of open issues and challenges in the scheduling of MapReduce to improve its performance. Copyright © 2015 John Wiley & Sons, Ltd.

Summary In recent years, many energy‐efficient algorithms and data dissemination protocols have been proposed for wireless sensor networks (WSNs). Because sensor nodes close to sink node have more traffic loads, they will quickly deplete their limited energy in practical implementation, and it will finally lead to energy hole and network partition problem. Adding sink mobility into sensor networks can bring in new opportunities to improve energy efficiency for WSNs. In this paper, we proposed our multiple mobile sink‐based routing algorithm for data dissemination to improve WSNs performance. Multiple mobile sinks will be utilized to collect the interested raw data. They will move back and forth along predetermined paths; one of which is the diameter of the circle, and the other two are fixed on arc lines. Mobile sinks will sojourn at some fixed points to collect raw data from relevant areas. Extensive simulation results show that our proposed algorithm can efficiently mitigate the hot spots problem and prolong the network lifetime of WSNs. Copyright © 2014 John Wiley & Sons, Ltd.

Healthcare domain is continuously growing with new knowledge emerged at different levels of clinical interest. At the same time, there is an increasing interest in the use of clinical decision support systems (CDSSs) to increase the healthcare quality and efficiency. Majorly the existing CDSSs are not designed to adapt scientific research in a well-established and automatic manner. Clinicians and researchers access the online resources on frequent basis for unmet questions during the course of patient care. They usually follow a dis-integrated approach to search for their required information from resources of their interest. Additionally, there is lack of defined mechanism to integrate the relevant knowledge for future use. To overcome the disintegrated and non-automatic approach, we introduce the concept of KnowledgeButton; a comprehensive model for evidence adaption from online credible knowledge sources in a well-defined and established manner. It saves the time of clinicians spend unnecessary in searching research evidence using disintegrated and manual mechanism. In this paper, we provide architecture design, workflows, and scenarios complemented with primary results. It covers walk-through from search query generation to evaluation of search results.

Clinical Decision Support Systems assist the physicians to make critical decisions during the diagnosis and treatment of the patients. CDSS envisions an extendable, shareable and reusable knowledge base to generate shareable guidelines and recommendations. Existing systems utilize HL7 standard Arden Syntax MLM and data model vMR schema for shareability and interoperability purpose. Understanding the Arden Syntax and vMR schema classes is tedious task for physicians, therefore an easy to use knowledge authoring environment is required to overcome the barrier of acquiring clinical knowledge. We are presenting Arden Syntax based authoring environment called Arden Syntax Studio that provides an easy to use interface to the physicians for creating shareable MLM without understanding the HL7 standard Arden Syntax. The created MLM can easily integrate with other healthcare systems by using HL7 standard data model vMR. The system hides the vMR layer from physicians by replacing vMR schema class's layer with corresponding understandable healthcare system's concepts. Therefore, the physician does not need to learn and understand complete Arden Syntax and vMR data model to create shareable rules in form of MLM. The system also provides the Intelli-sense functionality to enhance the knowledge rule creation process and reduce the possibilities of physician's errors. Furthermore, to enhance the interoperability feature our system capitalizes standard terminologies of the SNOMED CT concepts. The terminologies that are specific to organization, environment or region are handled with localized concepts in addition to SNOMED CT concepts.

Hadoop based Map Reduce (MR) has emerged as big data processing mechanism in terms of its data intensive applications. In data intensive systems, analysis and visualizations as a result of various algorithms can lead to differentiable and comparable results. Current implementations of MR facilitates to reuse the results of MR jobs in other MR jobs and to distribute the cloud resources among jobs. However, very little work is done in terms of using same data for multiple algorithms at the same time in a single job using either shared resources or dynamic resource allocation based on the data and scheduling of Map Reduce jobs. In this paper we propose a method to execute multiple algorithms on same data in HDFS concurrently and to use the same available resources by dynamically managing the task assignment and results aggregation. Our proposed approach reduces the execution time and supports multiple algorithms execution in parallel. In-Map/In-Reduce shows 200% decrease in execution time.

The rise of social media in couple of years has changed the general perspective of networking, socialization, and personalization. Use of data from social networks for different purposes, such as election prediction, sentimental analysis, marketing, communication, business, and education, is increasing day by day. Precise extraction of valuable information from short text messages posted on social media (Twitter) is a collaborative task. In this paper, we analyze tweets to classify data and sentiments from Twitter more precisely. The information from tweets are extracted using keyword based knowledge extraction. Moreover, the extracted knowledge is further enhanced using domain specific seed based enrichment technique. The proposed methodology facilitates the extraction of keywords, entities, synonyms, and parts of speech from tweets which are then used for tweets classification and sentimental analysis. The proposed system is tested on a collection of 40,000 tweets. The proposed methodology has performed better than the existing system in terms of tweets classification and sentiment analysis. By applying the Knowledge Enhancer and Synonym Binder module on the extracted information we have achieved increase in information gain in a range of 0.1% to 55%. The increase in information gain has enabled our proposed system to better summarize the twitter data for user sentiments regarding a keyword from a particular category.

MPI provides high performance parallel file access API called MPI-IO. ROMIO library implements MPI-IO specifications thus providing this facility to C and Fortran programmers. Similarly, object-oriented languages such as Java and C# have adapted MPI specifications and their implementations provide HPC facility to its programmers. These implementations, however, lack parallel file access capability which is very important for large-scale parallel applications. In this paper, we propose a Java based parallel file access API called MPJ-IO and describe its reference implementation. We describe design details and performance evaluation of this implementation. We use JNI calls in our code to utilize functions from ROMIO library. In addition, we highlight the reasons for using JNI calls in our code.

Mobile cloud computing aims at improving user experience through enhancing the ability of mobile applications by doing intensive tasks in the cloud. In this paper we consider an environment similar to a hybrid cloud in which the mobile device works as a private cloud. Given that the mobile phone has both limited processing resources and battery time, the proposed mobile application architecture has been designed with the capability of sending specified data/parameters to the cloud. This data is subsequently used for further processing/mining and visualization to assist in inferring further information through mapreduce. This information gives details about resource and battery consumption which will help in optimizing the relationship between the mobile device and cloud. It will also be beneficial to the optimization of the mobile application through the trends visualized in the cloud. In this paper we created an activity recognition health application as an example and helped the user about his health along with giving an insight into abnormal behavior and lifestyle trends.

Social media is a mean to connect people through information sharing. It has been helping to use in various domains like education, business and human resource management. Healthcare domain however can be considered as one of the most potential domain to benefit from the social media participation in its different sub-domains like clinical observation, medications, professional training, patient education, home care and personalize care. However, there has always been a debate about contents' trust worthiness. Nevertheless, social media is totally discouraged to be considered in healthcare due to fear of open and unauthenticated contents. With all these reservations, the importance of information shared through social networks can never be ignored. This paper investigates the role of social media in patient centric healthcare. It highlights the special and active contribution of the social technologies in the domain of Decision Support System (DSS). The outcomes of this work will results in improved architecture for Smart CDSS system which is under development at UC Lab (Ubiquitous Computing Research Lab Kyung Hee University, South Korea).

Cloud Computing has been adopted by health-care industry for their data's storage, manipulation, and secured sharing needs. However, cloud's distributed nature can be exploited to the use of scientific applications that are designed for health-care data evaluation. These scientific applications, such as, Medical Imaging, Gene and Protein annotation, Mapdrug therapy and Clinical Decision Support Systems (CDSS), require high-performance messaging libraries with minimum computational and communication overhead, and efficient resource utilization. The proposed High Performance Java Sockets (HPJS) encapsulates the needs of high-performance messaging of scientific applications for cloud platforms. HPJS effectively uses Java's socket implementation for high-performance inter-process communication. With single-copy protocol, thread re-usability and reduced communication overhead, HPJS can perform messaging twice as fast to conventional buffered-base communication libraries.

With the advent of smart technologies potential ideas have been emerged to facilitate human lives. Based on sensor technologies, smart homes concept is prevailing now a days that intends to bring tremendous changes in human lifestyle. The most prominent application is to equip the smart home with monitoring system that facilitate in managing care for elderly people. Elderly people with chronic disease need continuous care for managing their activities specially medications. The cost is increasing on care of elderly people and often needs sparing of family resource to take care during management of their activities and medications. This paper propose idea of Clinical Decision Support Service (CDSS) that provides guidelines and recommendation based on observed activities of patient. Our proposed CDSS service called Smart CDSS is deployed on platform that support various sensors and emotion recognition applications. The Smart CDSS knowledge base is currently supporting diabetes rules extracted from online resources and validated against recommendation from physician for 100 patients during their visits to local hospital. The Smart CDSS service allow interaction through standard base interfaces following HL7 vMR standard that allow seamless integration to underlying platform. Moreover, HL7 Arden Syntax is incorporated to scale up knowledge base for other diseases and allows sharing of clinician knowledge.

An important feature of the intelligent environments is that they monitor user activities and help them in making decisions based on their progress in activities. Many such environments have been designed previously for healthcare management. However, more than often users are reluctant to consider the feedback from the environment alone. So we consider expert's recommendations and social network of the user as additional entities that form part of the intelligent environment. However, unlike traditional approaches, they influence the user indirectly through various persuasion techniques. The main idea is to change the behavior of the user for improving healthcare management. In this article we identify salient features of our framework for healthcare self-management in intelligent environments that combines ubiquitous and social computing as persuasion media. The framework enables social interactions between the patients, doctors, and other users in their online social community through a web portal as well as through their smart phones. Both user's behavior and preferences are taken into account to help them in adopting healthy behavior. This is done by using different persuasion strategies created on the basis of the user's behavior model. As a case study we consider diabetes self-management in this article.

Modern HPC applications put forward significant I/O requirements. To deal with them, MPI provides the MPI-IO API for parallel file access. ROMIO library implements MPI-IO and provides efficient support for parallel I/O in C and Fortran based applications. On the other hand, Java based MPI-like libraries such as MPJ Express and F-MPJ have emerged but they lack parallel I/O support. Little research has been done to provide Java based ROMIO-like libraries due to the non-availability of MPI-IO-like API for the Java language. In this paper, we take the first step towards the development of parallel I/O API in Java by evaluating the newly introduced Java NIO API versus the legacy Java I/O API. We propose two simple approaches for performing parallel file I/O using NIO and evaluate them on two different computational platforms. The implementation of proposed approaches exploits the view buffers concept of NIO API to perform efficient array based file I/O operations from multiple processes. We report encouraging speedups and suggest that design of a parallel I/O API in Java should be based on the NIO API.

Automatic foreground object segmentation is a fascinating, a demanding research area, and an exigent problem in biomedical applications. Existing works cannot segment concave objects and completely dependent on initial curve that is initialized manually by the users, and must be closer to the object. Due to these limitations, most of them were considered as semi-automatic approaches. In this paper, we incorporated active contours (level-set) based on Bhattacharya distance to the Chan and Vese energy functional such that are not only minimized the differences within each region but also maximized the distance between the two regions as well. Compared with active contour snake, the proposed model gave more accurate results that segment the foreground objects automatically.

This paper presents an approach for scheduling U-Life care applications in the cloud computing environment based on virtual resources to support real-time services and to improve user Quality of Service (QoS) requirements. We design and develop an architecture called ULC3 (Ubiquitous Life Care Cloud Computing) that uses virtual resources provided by cloud computing to schedule U-Life care applications. The ULC3 is based on the concepts of cloud computing and wireless sensor networks. The architecture is very important and necessary to support create virtual clusters dynamically, deploys the required number of virtual machines (VMs) in potential computing resources to meet the application requirements, and to configure with the required software execution environment. Thus, the system can improve computation time, guarantee the QoS, and support real-time services. Finally, the results from the execute applications are provided to the end-users as a service.

The quick-start mechanism has been presented as a way for transport control protocols to efficiently use available bandwidth in under-utilized networks. Although studies of performance of quick-start have been carried out, its performance in low latency networks has not been well investigated. In this paper, we show that the benefits of quick-start in low latency networks are minimal. We also show that the improvement in network utilization is much lower than in networks with higher latencies. Our future work in this field will undoubtedly lead to new algorithms and improvements to existing ones.

This paper presents a Secured Wireless Sensor Network-integrated Cloud computing for u-Life Care (SC3). SC3 monitors human health, activities, and shares information among doctors, care-givers, clinics, and pharmacies in the Cloud, so that users can have better care with low cost. SC3 incorporates various technologies with novel ideas including; sensor networks, Cloud computing security, and activities recognition.

Trust-aware recommender system (TARS) suggests the worthwhile information to the users on the basis of trust. Existing works of TARS suffers from the problem that they need extra user efforts to label the trust statements. The authors propose a novel model named iTARS to improve the existing TARS by using the implicit trust networks: instead of using the effort-consuming explicit trust, the easy available user similarity information is used to generate the implicit trusts for TARS. Further analysis shows that the implicit trust network has the small-world topology, which is independent of its dynamics. The rating prediction mechanism of iTARS is based on the small worldness of the implicit trust network: the authors set the maximum trust propagation distance of iTARS approximately equals the average path length of the trust network's corresponding random network. Experimental results show that with the same computational complexity, iTARS is able to improve the existing TARS works with higher rating prediction accuracy and slightly worse rating prediction coverage.

Physical-activity recognition via wearable sensors can provide valuable information regarding an individual's degree of functional ability and lifestyle. In this paper, we present an accelerometer sensor-based approach for human-activity recognition. Our proposed recognition method uses a hierarchical scheme. At the lower level, the state to which an activity belongs, i.e., static, transition, or dynamic, is recognized by means of statistical signal features and artificial-neural nets (ANNs). The upper level recognition uses the autoregressive (AR) modeling of the acceleration signals, thus, incorporating the derived AR-coefficients along with the signal-magnitude area and tilt angle to form an augmented-feature vector. The resulting feature vector is further processed by the linear-discriminant analysis and ANNs to recognize a particular human activity. Our proposed activity-recognition method recognizes three states and 15 activities with an average accuracy of 97.9% using only a single triaxial accelerometer attached to the subject's chest.

Abstract Cognitive radio (CR) has been considered a key technology to enhance overall spectrum utilization by opportunistic transmissions in CR transmitter–receiver link(s). However, CRs must form a cognitive radio network (CRN) so that the messages can be forwarded from source to destination, on top of a number of opportunistic links from co‐existing multi‐radio systems. Unfortunately, appropriate routing in CRN of coexisting multi‐radio systems remains an open problem. We explore the fundamental behaviors of CR links to conclude three major challenges, and thus decompose general CRN into cognitive radio relay network (CRRN), CR uplink relay network, CR downlink relay network, and tunneling (or core) network. Due to extremely dynamic nature of CR links, traditional routing to maintain end‐to‐end routing table for ad hoc networks is not feasible. We locally build up one‐step forward table at each CR to proceed based on spectrum sensing to determine trend of paths from source to destination, while primary systems (PSs) follow original ways to forward packets like tunneling. From simulations over ad hoc with infrastructure network topology and random network topology, we demonstrate such simple routing concept known as CRN local on‐demand (CLOD) routing to be realistic at reasonable routing delay to route packets through. Copyright © 2009 John Wiley & Sons, Ltd.

Key establishment plays a central role in authentication and encryption in wireless sensor networks, especially when they are mainly deployed in hostile environments. Because of the strict constraints in power, processing and storage, designing an efficient key establishment protocol is not a trivial task. Compared with public key cryptography, symmetric key cryptographic with key predistribution mechanism is more suitable for large-scale wireless sensor networks. Most of previous solutions have some issues on performance and security capabilities. In this paper, we propose a novel key predistribution model using pre-deployment knowledge and random values in pairwise key generation to take advantage in terms of network connectivity, memory cost, energy for transmission and strong resilience against node capture attacks.

Recent research on ubiquitous computing has introduced a new concept of activity-based computing as a way of thinking about supporting human activities in ubiquitous computing environment. Existing access control approaches such as RBAC, became inappropriate to support this concept because they do not consider human activities. In this paper, we propose Activity-Oriented Access Control (AOAC) model, aiming to support user's activity in ubiquitous environments. We have designed and implemented our initial AOAC system. We also built up a simple scenario in order to illustrate how it supports user activities. The results have shown that AOAC meets our objectives. Also, AOAC it takes approximately 0.26 second to give a response which proves that AOAC is suitable to work in real-time environments.

For a learning task with a huge number of training instances, we sample some informative/important instances, which are then used for learning. Obtaining accurately labeling data is always difficult thus noise detection is required to filter out noises from sampled instances since the noises will degrade the learning performance. In this work, we propose to utilize unsampled instances to improve the performance of noise detection in sampled instances. Empirical study validates our idea that refined classifier can be achieved from noisy sampled instances by utilizing unsampled instances.

Knowledge constantly grows in scientific discourse and is revised over time by domain experts. The body of knowledge will get structured and refined as the communities of practice concerned with the field of knowledge develop a deeper understanding of issues. The knowledge model, as a result evolves to a new state to accommodate the new knowledge. Keeping trail of these changes in semantically rich and formally sound mechanism, has pragmatic advantages for providing the undo and redo facility and recover to a previous state of the knowledge body (i.e. ontology). In this research, we have developed and tested comprehensive methodological framework for change tracer. The ontology changes are captured and then stored in change history log (CHL) in conformance to change history ontology (CHO). The CHL is later used for reverting ontology to a previous consistent state and visualization of change effects on ontology. The system is compared with ChangesTab of Protege, a comprehensive evaluation of the accuracy of roll-back and roll-forward algorithm has been conducted over documentation ontology. The system is also tested over a standard dataset of OMV and high accuracy results are observed for both roll-back and roll-forward algorithms.

Cloud computing allows users to use only a Web browser to receive computing services via the Internet. Users only need to pay for the services they actually use. It appears that a wide adoption of cloud computing in the foreseeable future is inevitable, and its adoption will bring about a sea change in the pricing and distribution practices for both software and hardware. There are, however, various issues that will impede adoption of cloud computing. Most of them can be solved. We discuss the status of cloud computing today and various adoption issues. We also provide a market prognosis.

In this paper, we present a new approach to solving the distributed facility location problem using the recent modeling and computational methodology of factor graph and message-passing. We first formulate the problem as finding a valid network configuration that minimizes the overall cost. We then represent the problem using a factor graph, and derive simplified, localized, broadcast-based message-passing rules which can elect a near-optimal set of facility nodes in a few iterations. Simulation results for small-world network topologies show that the algorithm is able to achieve good convergence rate and approximation ratio, and scalable to the network size.

Cloud computing allows users to use only a Web browser to receive computing services via the Internet. Users only need to pay for the services they actually use. It appears that a wide adoption of cloud computing in the foreseeable future is inevitable, and its adoption will bring about a sea change in the pricing and distribution practices for both software and hardware. There are, however, various issues that will impede adoption of cloud computing. Most of them can be solved. We discuss the status of cloud computing today and various adoption issues. We also provide a market prognosis.

Traditional trust management schemes developed for wired and wireless ad hoc networks are not well suited for sensor networks due to their higher consumption of resources such as memory and power. In this work, we propose a new lightweight group-based trust management scheme (GTMS) for wireless sensor networks, which employs clustering. Our approach reduces the cost of trust evaluation. Also, theoretical as well as simulation results show that our scheme demands less memory, energy, and communication overheads as compared to the current state-of-the-art trust management schemes and it is more suitable for large-scale sensor networks. Furthermore, GTMS also enables us to detect and prevent malicious, selfish, and faulty nodes.

Full network level privacy spectrum comprises of identity, route, location and data privacy. Existing privacy schemes of wireless sensor networks only provide partial network level privacy. Providing full network level privacy is a critical and challenging problem due to the constraints imposed by the sensor nodes, sensor networks and QoS issues. In this paper, we propose full network level privacy solution that addresses this problem. This solution comprises of Identity, Route and Location (IRL) privacy algorithm and data privacy mechanism, that collectively provides protection against privacy disclosure attacks such as eavesdropping and hop-by-hop trace back attacks.

The performance of supervised learning could be improved when valuable data are selected for training. In this paper, we proposed three data selection methods based on fuzzy C-means algorithm. They are: center-based selection, border-based selection and bin-based selection. In center-based selection, the data with high degree of membership in each cluster are selected for training. In border-based selection, the data around the borders between clusters are selected. In bin-based selection, the data in each cluster are sorted based on their membership degrees. Then for each cluster, the sorted data are divided into bins. Finally, there is one data selected from each bin for training. The effects of them are empirically studied on a set of UCI data sets. Experimental results indicate that bin-based selection could effectively improve the performance of learning compared to randomly selecting training samples.

This paper proposes a novel method for data editing. The goal of data editing in instance-based learning is to remove instances from a training set in order to increase the accuracy of a classifier. To the best of our knowledge, although many diverse data editing methods have been proposed, this is the first work which uses semi-supervised learning for data editing. Wilson editing is a popular data editing technique and we implement our approach based on it. Our approach is termed semi-supervised nearest neighbor editing (SSNNE). Our empirical evaluation using 12 UCI datasets shows that SSNNE outperforms KNN and Wilson editing in terms of generalization ability.

Bardram introduced a new concept of activity-based computing as a way of thinking about supporting human activities in ubiquitous environments. In such environments where users are using a multitude of heterogeneous computing devices, the need for supporting users at the activity level becomes essential. However, without considering basic security issues, it could be rife with vulnerabilities. Security services, like authentication and access control, have to not only guarantee security, privacy, and confidentiality for ubiquitous computing resources, but also support user activities equipped with various devices. In this paper, we present an activity-based security scheme. The proposed scheme aims to enhance security services on mobile devices and facilitate user activities. We also integrate off-the-shell security services like MD5, TEA, Diffie-Hellman key agreement protocol so that it makes the scheme more robust and practically usable. The implementation and sample scenario have shown the requirement satisfactory of the scheme.

As the cornerstone of effective patient-physician relationships in the traditional healthcare infrastructures, trust faces new opportunities and challenges in the ubiquitous healthcare. Ubiquitous healthcare enables the agents acquire more information on trust evaluation through effectively resource sharing. Yet ubiquitous healthcare also lays the agents in a more dynamic and uncertainty environment for the trust evaluations. This paper contributes to develop a distributed trust management for the ubiquitous healthcare. Our trust management infrastructure is responsible for evaluating the trust value and assigning access rights based on the trust value. Based on each agent's confidence of its personal experience on other agents, three naive Bayes classifier based algorithms are introduced for the trust evaluation: the robust experience algorithm, the weak experience algorithm and the no experience algorithm. The simulation results show the feasibility and effectiveness of our trust management in the ubiquitous healthcare.

Recent technological advances in wireless sensor networking have opened up new opportunities in healthcare systems. Future medical systems are expected to benefit the most in such areas as in-home assistance, smart nursing homes, and clinical trial. However the network system including body sensor networks and environmental sensor networks are normally comprised of energy constrained nodes. Furthermore, communication interference between multi mode nodes in such a dynamic system is also a challenge. This limitation has led to the crucial need for energy and mobility aware protocols to produce an efficient network. In this paper, we propose an energy and mobility aware multipath routing scheme for sensor networks in a smart homecare application. The remaining battery capacity, distance to the gateway, mobility and queue size of candidate sensor nodes in the local communication range are taken into consideration for next hop relay node selection, and Analytical Hierarchy Process (AHP) is applied for decision making. Simulation results show that this scheme can extend the network lifetime and reduce the packet loss rate and link failure rate since the mobility and buffer capacity is considered.

Important applications of Wireless Ad Hoc Networks make them very attractive to attackers, therefore more research is required to guarantee the security for Wireless Ad Hoc Networks. In this paper, we proposed a transmission time based mechanism (TTM) to detect wormhole attacks - one of the most popular & serious attacks in Wireless Ad Hoc Networks. TTM detects wormhole attacks during route setup procedure by computing transmission time between every two successive nodes along the established path. Wormhole is identified base on the fact that transmission time between two fake neighbors created by wormhole is considerably higher than that between two real neighbors which are within radio range of each other. TTM has good performance, little overhead and no special hardware is required.

Networks make them very attractive to attackers, therefore more research is required to guarantee the security for Wireless Ad Hoc Networks. In this paper, we proposed a transmission time based mechanism (TTM) to detect wormhole attacks – one of the most popular & serious attacks in Wireless Ad Hoc Networks. TTM detects wormhole attacks during route setup procedure by computing transmission time between every two successive nodes along the established path. Wormhole is identified base on the fact that transmission time between two fake neighbors created by wormhole is considerably higher than that between two real neighbors which are within radio range of each other. TTM has good performance, little overhead and no special hardware required. TTM is designed specifically for Ad Hoc On-Demand Vector Routing Protocol (AODV) but it can be extended to work with other routing protocols.

Due to mobility, energy limitations, and unreliable wireless channels, applications running on mobile devices suffer from faults such as temporary disconnection and data loss. We, therefore, need a fault tolerance mechanism to guarantee their smooth working and performance. In this paper, we present a novel proxy-based uncoordinated checkpointing scheme with pessimistic message logging for efficient fault recovery in mobile Grid system. Simulation results show that this scheme is reliable, efficient and, at the sametime, consumes less network traffic.

The provision of embedding neural networks into software applications can enable variety of artificial intelligence systems for individual users as well as organizations. Previously, software implementation of neural networks remained limited to only simulations or application specific solutions. Tightly coupled solutions end up in monolithic systems and non reusable programming efforts. We adapt component based software engineering approach to effortlessly integrate neural network models into AI systems in an application independent way. As proof of concept, this paper presents componentization of three famous neural network models i) multi layer perceptron ii) learning vector quantization and iii) adaptive resonance theory family of networks.

In this work we present a new technique of facial-image retrieval using constrained independent component analysis (cICA). We have employed cICA for the online extraction of those independent components from the entire database which bear some similarity to the query-images. Instead of using any offline learning mechanism or feature extraction technique, our system works completely online. It can cater for queries formulated from both single and multiple examples, for achieving higher accuracy. For compound queries, instead of treating each query-image independently, the system is capable of finding images similar not only to the individual query-images, but also to their different combinations

Received signal strength (RSS) based location systems are potential candidates to enable indoor location aware services due to pervasively available wireless local area networks and hand held devices. Intrinsically RSS based positioning is a multi-class pattern recognition problem. Previous researches have shown that Visibility Matrix based approach of modular classifiers improves location accuracy but costs longer periods of training and testing in development life cycle. We present a context-aware fuzzy ArtMap neural network that provides competitive location accuracy in comparison with modular approach while leveraging online and incremental learning capabilities to location system development life cycle.

This paper presents novel hybrid architecture of control system of mobile robot oriented on navigation tasks and based on hybrid neural network MLP-ART2. Using of this model allows movement to target avoiding the obstacles almost without assistance of human operator after just small learning.

Reputation system is a way to maintain trust in dynamic environments by collecting, distributing and aggregating feedbacks about the service providers' past behaviors. Most existing reputation systems assume that raters evaluate the ratee by means of numerical values. However, raters sometimes cannot express their judgments with exact numerical values, especially when the raters have uncertain or ambiguous opinions on the ratee. Our paper introduces a novel reputation system based on the methodology of Computing with Words (CW), in which the ratings and reputations of computation are words and propositions drawn from a natural language instead of numerical values. Our reputation system has a sound mathematical basis. At the same time, it is convenient for the raters to express their judgments and simple for the participants to understand the integrated reputation.

The technology of multimedia content adaptation based upon the location of a target device can become the long expected killer application of ubiquitous computing. Easy to develop, lightweight, and robust location estimation is the core component of this technology. Until now, location estimation technology remains restricted to highly sophisticated hardware and networking infrastructure where semantics of the location information are defined and controlled by service providers. We aim to lower the technical and infrastructure barriers to allow general users to define and develop the semantically meaningful location systems. This paper presents a simple location estimation method to build radio beacon based location systems in the indoor environments. It employs an realtime learning approach which requires zero prior knowledge. The salient features of our method are low memory requirements and simple computations which make it desirable for location-aware multimedia systems functioning in distributed client-server settings as well as privacy sensitive applications residing on stand alone devices.

Compared to all emerging issues, privacy is probably the most prominent concern when it comes to judging the effects of a wide spread deployment of ubiquitous computing. On one hand, service providers want to authenticate legitimate users and make sure they are accessing their authorized services in a legal way. On the other hand, users prefer not to expose any sensitive information to anybody. They want to have complete control on their personal data, without being tracked down for wherever they are, whenever and whatever they do. In this paper, we introduce an anonymous identification authentication and access control scheme to secure interactions between users and services in ubiquitous environments. The scheme uses anonymous user ID, sensitive data sharing method, and account management to provide a lightweight authentication while keeping users anonymously interacting with the services in a secure and flexible way.

Activity recognition is a hot topic in context-aware computing. In activity recognition, machine learning techniques have been widely applied to learn the activity models from labeled activity samples. Since labeling samples requires human's efforts, most existing research in activity recognition focus on refining learning techniques to utilize the costly labeled samples as effectively as possible. However, few of them consider using the costless unlabeled samples to boost learning performance. In this work, we propose a novel semi-supervised learning algorithm named En-Co-training to make use of the unlabeled samples. Our algorithm extends the co- training paradigm by using ensemble method. Experimental results show that En-Co-training is able to utilize the available unlabeled samples to enhance the performance of activity learning with a limited number of labeled samples.

Hospital work is characterized by the need to manage multiple activities simultaneously, constant local mobility, frequently interruptions, and intense collaboration and communication. Hospital employees must handle a large amount of data that is often tied to specific work activities. This calls for a proper access control model. In this paper, we propose a novel approach, activity-based access control model (ACM). Unlike conventional approaches which exploit user identity/role information, ACM leverages user's activities to determine the access permissions for that user. In ACM, a user is assigned to perform a number of actions if s/he poses a set of satisfactory attributes. Access permissions to hospital information are granted according to user's actions. By doing this, ACM contributes a number of advantages over conventional models: (1) facilitates user's work; (2) reduces complexity and cost of access management. Though the design of ACM first aims to support clinical works in hospitals, it can be applied in other activity-centered environments.

Location Awareness is key capability of Context-Aware Ubiquitous environments. Received Signal Strength (RSS) based localization is increasingly popular choice especially for in-building scenarios after pervasive adoption of IEEE 802.11 Wireless LAN. Fundamental requirement of such localization systems is to estimate location from RSS at a particular location. Multipath propagation effects make RSS to fluctuate in unpredictable manner, introducing uncertainty in location estimation. Moreover, in real life situations RSS values are not available at some locations all the time making the problem more difficult. We employ Modular Multi Layer Perceptron (MMLP) approach to effectively reduce the uncertainty in location estimation system. It provides better location estimation results than other approaches and systematically caters for unavailable signals at estimation time.

In this paper, we propose an approach to collaboratively track motion of a moving target in a wide area utilizing camera-equipped visual sensor networks, which are expected to play an essential role in a variety of applications such as surveillance and monitoring. A genetic fitting method for efficient contour extraction is used as inter-scene approach to detect and track the target. We also considered the existence of faulty sensors in the network which deteriorate the difficulty of target tracking problem, and proposed a robust sensor collaboration method. The experimental results have shown that the proposed target tracking approach produces very successful target tracking compared with the existing method especially in case that the target is adjacent to neighboring objects of background

Traditional authentication and access control are no longer suitable for ubiquitous computing paradigm. They are only effective if the system knows in advance which users are going to access and what their access rights are. Therefore, it calls for a novel security model. In this paper, we outline major security problems in ubiquitous computing and propose a new architecture, TBSI (Trust-based Security Infrastructure). In TBSI, trust and risk management plays a key role to support authentication and authorization to unknown users. Meanwhile, intrusion detection and home firewall are also integrated to make TBSI more robust. This paper is an extension of our previous work to give more detailed description and enhancement of the architecture. TBSI is on-going research project to support our context-aware middleware CAMUS.

Sensor network security solutions that have been proposed so far are mostly built on the assumption of a trusted environment, which is not very realistic so we need trust management before deploying any other security solution. Traditional trust management schemes that have been developed for wired and wireless ad-hoc networks are not suitable for wireless sensor networks because of higher consumption of resources such as memory and power. In this paper, we propose a novel lightweight group based trust management scheme (GTMS) for distributed wireless sensor networks in which the whole group will get a single trust value. Instead of using completely centralized or distributed trust management schemes, GTMS uses hybrid trust management approach that helps in keeping minimum resource utilization at the sensor nodes.

A roaming user in ubiquitous environment should have access to different services anywhere anytime. In an infrastructure based smart environment, user acquires this facility from the middleware. Syntax based service discovery has proven to be inadequate for flexible interaction between user and the middleware; context based semantic matching is necessary. This paper shows the use of ontology to facilitate the semantic discovery of services in a ubiquitous middleware named Context Aware Middleware for Ubiquitous System (CAMUS).

In this paper, we study fault tolerant routing and broadcasting in interconnection networks based on de Bruijn graph (dBG) for constructing large scale multiprocessors networks. Our paper presents a new approach to provide fault tolerance routing and broadcasting which haven't been investigated. The proposed approach is based on multi level discrete set concept in order to find a fault free shortest path among several paths provided. In the proposed fault tolerant broadcasting, we can achieve k (network diameter) as maximum time step to finish broadcast process and there is no overhead in the broadcast message.

Ubiquitous computing foresees a massively networked world supporting a population of diverse but cooperating mobile devices where trust relationships between entities are uncertain. Though there have been lots of effort focusing on trust for ubiquitous systems, they did not attach enough importance to uncertainty in their model. On the other hand, most of the works draw a general picture without a detailed computational model. In this paper, we present a trust model based on the vectors of trust values of different entities. The evaluation of trust depends upon the recommendation of peer entities common to the interacting entities. These recommendations are weighted according to the number and time of past interactions. Furthermore we present a method of handling false recommendations without introducing significant computational burden. The model can calculate trust between two entities in situations both in which there is past experience among the interacting entities and in which the two entities are communicating for the first time. Several tuning parameters are suggested which can be adjusted to meet the security requirement of a ubiquitous system.

Location based services are becoming essential feature of context-awareness in ubiquitous computing. Reflective distribute component programming model is proposed to systematically provide location to location based services (LBS). We integrate distributed component technology and reflection to develop localization capability as middleware service. Concept of meta object protocols (Reflection) is used in different way than traditional reflective mechanisms. It deals with the state of the component that is not inside the component rather resides outside of it, namely extrinsic. This component model provides the basis for middleware architecture to support location providing service at design, implementation and run time. We describe the methodology we used to build location-awareness as middleware service based upon our reflective component model.

Context aware middleware infrastructures have traditionally been implemented with a modular approach to allow different components to work cooperatively and supply context synthesis and provision services. In this paper, we discuss the important requirements that arise when such a middleware is deployed in a distributed environment and present the design and implementation of context aware middleware for ubiquitous systems (CAMUS) with which the authors have attempted to meet those requirements. Issues related to distributed coordination within the middleware in terms of component discovery and management and multiple context domains are also discussed.

Context-aware middleware encompasses uniform abstractions and reliable services for common operations, supports for most of the tasks involved in dealing with context, and thus simplifying the development of context-aware applications. In this paper, we address some key issues of a middleware for context-aware ubiquitous computing, ranging from design considerations of a unified sensing framework, formal modeling and representation of the real world, pluggable reasoning engines for high-level contexts, and context delivery-runtime service composition mechanisms. Our implementation experience indicates that a comprehensive approach throughout the system layers results in a flexible and reusable middleware architecture.

Large-scale distributed systems, such as ubiquitous computing environments, require a service delivery mechanism in order to keep track of the vast set of services offered and make them available to interested clients. The amount of services and clients, their context, and loose coupling between them makes service delivery in ubiquitous environments different from other systems. This paper presents a solution to overcome these issues by utilizing the underlying ontology and semantics for service lookup. Access control over context data is also considered by specifying dynamic policies at the system and service level.

Many efficient data dissemination protocols for mobile sinks in large scale sensor networks are currently under developed by researchers. In this paper we propose CODE, a coordination-based data dissemination protocol for wireless sensor networks, CODE relies on grid structure and GAF protocol to achieve better energy consumption by establishing an efficient data dissemination path and turning off unnecessary nodes. Our simulation results show that CODE achieves more energy efficient and longer networks life time compared with other approaches while still handling efficient data delivery to mobile sinks.

The major drawback of the slack-stealing based schedulings for aperiodic requests is a high computational complexity to calculate the slack which in consequence makes them not be practical. In this paper, we present a soft-aperiodic task scheduling algorithm, called Alternative Priority Scheduling (APS), which has a simple slack calculation method in dynamic priority systems. The proposed algorithm has extended the EDF-CTI (Earliest Deadline First-Critical Task Indicating) Algorithm developed by the authors. The APS algorithm references the off-line built CTI table and chooses either an EDF or a CEF (Critical Execution time First) algorithm alternatively at run-time. This paper also demonstrates the optimality of the APS algorithm. Our simulation study shows that the APS algorithm, in most cases, is slightly better than the EDF-CTI algorithm and the other soft-aperiodic schedulings in terms of the short response time of aperiodic requests, and considerably improves the previous algorithms in a high workload.