Search

Nowadays, the explosion of CCTV cameras has resulted in an increasing demand for distributed solutions to efficiently process the vast volume of video data. Otherwise, the use of surveillance when people are being watched remotely and recorded continuously has raised a significant threat to visual privacy. Using existing systems cannot prevent any party from exploiting unwanted personal data of others. In this paper, we develop an intelligent surveillance system with integrated privacy protection, where it is built on the top of big data tools, i.e., Kafka and Spark Streaming. To protect individual privacy, we propose a privacy-preserving solution based on effective face recognition and tracking mechanisms. Particularly, we associate body pose with face to reduce privacy leaks across video frames. The body pose is also exploited to infer person-centric information like human activities. Extensive experiments conducted on benchmark datasets further demonstrate the efficiency of our system for various vision tasks.

Graph sparsification is a useful approach for mining, analyzing, and visualizing large graphs. It simplifies the structure of a graph by pruning some of the edges while preserving the nodes. One well-known edge-removal technique is determination of a single shortest path between any pair of nodes to maintain the overall connectivity of the graph and remove the rest of the edges. Recently, a graph sparsification approach has been proposed that classifies the edges for removal according to metricity. Considering this classification system, a three-phase algorithm has been presented to prune the edges of each identified classification. The main bottleneck of the existing approach is the computational cost required for all three phases. To overcome these problems, we propose a new concept that generalizes the metricity of the edges to a uniform and homogeneous level. We also propose a new algorithm: edge metricity–based faster graph sparsification (EM-FGS), which utilizes edge weight as a constraint for effective pruning of the search space for shortest path computations. Our proposed approach is effective because we use the edge weight, which is a freely available intrinsic property of a graph, for edge-metricity identification. Our proposed approach improves the time complexity of existing methods from O(m ∗ n) + O(U3) to O(m ∗ log(n)). We evaluated our approach on publicly available real and synthetic graphs and improved the sequential algorithm’s execution time by a minimum of 125 times and the parallel algorithm’s runtime speed performance by 13 times while providing the same degree of accuracy.

With the rise of web 2.0, a huge amount of unstructured data has been generated on regular basis in the form of comments, opinions, etc. This unstructured data contains useful information and can play a significant role in business decision making. In this context, sentiment analysis (SA) is an active research area and has recently attracted the attention of the research community. The aim of SA is to classify the user-generated content into positive and negative class. State-of-the-art techniques for sentiment classification relies on the traditional bag-of-words approaches. Such approaches can be advantageous in terms of simplicity but completely ignore the semantics aspects, the order between words, and also leads to the curse of dimensionality. Researchers have also proposed semantic-based SA techniques in conjunction with word-order employing high order n-grams, part-of-speech (POS) patterns, and dependency relation features. But can every word or phrase of high order n-grams, POS patterns or dependency relation features represent sentiment clue? If incorporated, then what about the dimensionality? In order to tackle and investigate such issues, in this paper, we propose a novel POS and n-gram based ensemble method for SA while considering semantics, sentiment clue, and order between words called EnSWF which is a four phase process. Our main contributions are four-fold (a) Appropriate Feature Extraction: we investigate and validate extracting various appropriate features for sentiment classification. (b) Dimensionality Reduction: We decrease the dimensionality of feature space by selecting the subset of most meaningful and effective features. (c) Ensemble Model: We propose an ensemble learning method for both filter based features selection and classification using simple majority voting technique. (d) Practicality: we authenticate our claim while applying our model on benchmark datasets. We also show that EnSWF out-perform existing techniques in terms of classification accuracy and reduce high dimensional feature space.

In this paper, we carry out an experimental analytics to show how much perfectly the conceptual mining framework is operable on re-discovering workflow process patterns and their enacted proportions from the workflow enactment event histories logged in a format of XES standardized schema. In principle, the framework must be able to properly handle all the workflow process patterns based upon the four types of control-flow primitives such as linear (sequential), disjunctive (selective), conjunctive (parallel), and loop (iterative) process patterns. The paper focuses on implementing an algorithmic mining framework only for discovering all the process patterns and their enacted proportions. To prove the functional correctness of the framework, we carry out an experimental mining and analytics on the real workflow instance enactment event histories of 10,000 workcases, and we finally visualize the mining and analytic artifacts and describe the implications of the results of the experiment.

Graphs are widely used in various applications, and their size is becoming larger over the passage of time. It is necessary to reduce their size to minimize main memory needs and to save the storage space on disk. For these purposes, graph summarization and compression approaches have been studied in various existing studies to reduce the size of a large graph. Graph summarization aggregates nodes having similar structural properties to represent a graph with reduced main memory requirements. Whereas graph compression applies various encoding techniques so that the resultant graph needs lesser storage space on disk. Considering usefulness of both the paradigms, we propose to obtain best of the both worlds by combining summarization and compression approaches. Hence, we present a greedy-based algorithm that greatly reduces the size of a large graph by applying both the compression and summarization. We also propose a novel cost model for calculating the compression ratio considering both the compression and summarization strategies. The algorithm uses the proposed cost model to determine whether to perform one or both of them in every iteration. Through comprehensive experiments on real-world datasets, we show that our proposed algorithm achieves a better compression ratio than only applying summarization approaches by up to 16%.

Background: Verifying the exact location of talocalcaneal (TC) coalition is important for surgery, but the complicated anatomy of the subtalar joint makes it difficult to visualize on radiographs. No study has used computed tomography (CT) or magnetic resonance imaging (MRI) to verify the radiological characteristics of TC coalition or those of different facet coalitions. Therefore, this study verified the radiological findings used to identify TC coalitions and those of different facet coalitions using CT and MRI. Materials and Methods: Plain with/without weight bearing anteroposterior and lateral radiographs, CT, and MRI of 43 feet in 39 patients with TC coalitions were reviewed retrospectively. CT or MRI was used to verify the location of the TC coalition. Secondary signs for the presence of a coalition in the anteroposterior and lateral plain radiographs, including talar beak, humpback sign, duck-face sign, and typical or deformed C-sign, were evaluated. Three independent observers evaluated the radiographs twice at 6-week intervals to determine intraobserver reliability. They examined the radiographs for the secondary signs, listed above, and coalition involved facets. Results: The average rates from both assessments were as follows: Middle facet 5%, middle and posterior facets 27%, and posterior facet 68%. The deformed C-sign is more prevalent in posterior facet coalitions. The posterior facet has the highest prevalence of involvement in TC coalitions, and the deformed C-sign and duck-face sign have high correlations with TC coalitions in the posterior subtalar facet. Conclusion: A posterior facet is the most prevalent for TC coalition, and the C-sign is useful for determining all types of TC coalition.

Background We hypothesized that calcaneal reconstruction can relieve chronic pain due to calcaneal malunion. We report the mid-term follow-up results of calcaneal reconstruction for calcaneal malunion. Methods We reviewed the records of 10 male patients (10 ft) who underwent calcaneal reconstruction for calcaneal malunion between January 2009 and July 2014 at the mid-term follow-up. Talocalcaneal height and angle, calcaneal pitch, calcaneal width, Böhler angle, Stephens classification, and Zwipp classification were evaluated by three orthopedic doctors at each visit (pre-reconstruction, post-reconstruction, and at the last follow-up). Results The mean follow-up period was 67.1 months (range, 48–101 months). The sites of pain before reconstruction were lateral aspect (4 patients), plantar aspect (3 patients), diffuse pain (2 patients), and anterior aspect (1 patient). There was a significant difference in talocalcaneal height, talocalcaneal angle, calcaneal pitch, calcaneal width, and Böhler angle before and after reconstruction (p

Graph summarization is valuable approach to analyze various real life phenomenon, like communities, influential nodes, and information flow in a big graph. To summarize a graph, nodes having similar neighbors are merged into super nodes and their corresponding edges are compressed into super edges. Existing methods find similar nodes either by nodes ordering or perform pairwise similarity computations. Compression-by-node ordering approaches are scalable but provide lesser compression due to exhaustive similarity computations of their counterparts. In this paper, we propose a novel set-based summarization approach that directly summarizes naturally occurring sets of similar nodes in a graph. Our approach is scalable since we avoid explicit similarity computations with non-similar nodes and merge sets of nodes in each iteration. Similarly, we provide good compression ratio as each set consists of highly similar nodes. To locate sets of similar nodes, we find candidate sets of similar nodes by using locality sensitive hashing. However, member nodes of every candidate set have varying similarities with each other. Therefore, we propose a heuristic based on similarity among degrees of candidate nodes, and a parameter-free pruning technique to effectively identify subset of highly similar nodes from candidate nodes. Through experiments on real world graphs, our approach requires lesser execution time than pairwise graph summarization, with margin of an order of magnitude in graphs containing nodes with highly diverse neighborhood, and produces summary at similar accuracy. Similarly, we observe comparable scalability against the compression-by-node ordering method, while providing better compression ratio.

An abundance of structural information has resulted in non-trivial graph traversals. Shortcut construction is among the utilized techniques implemented for efficient shortest path (SP) traversals on graphs. However, shortcut construction, being a computationally intensive task, required to be exclusive and offline, often produces unnecessary auxiliary data, i.e., shortcuts. Medium to large-scale graphs can take minutes to hours of computation time depending upon the utilization of computational resources and complexity of shortcut construction algorithms. In addition, the branching factor during SP expansions greatly increases due to excessive shortcuts. These factors make repeated SP queries unsuitable for graph mining tasks. This paper presents Shortest Path Overlapped Region (SPORE), a performance-based initiative that improves the shortcut construction performance by exploiting SP overlapped regions. Path overlapping has been overlooked by shortcut construction systems. SPORE takes advantage of this opportunity and provides a solution by constructing auxiliary shortcuts incrementally, using SP trees during traversals, instead of an exclusive step. SPORE is exposed to a graph clustering task, which requires extensive graph traversals to group similar vertices together, for realistic implications. We further suggest an optimization strategy to accelerate the performance of the clustering process using confined subgraph traversals. A performance evaluation of SPORE on real and synthetic graphs reveals an execution time gain of up to 40 %, having an order of magnitude fewer shortcuts over the SegTable approach. Leveraging the SPORE with multiple SP computations consistently reduces the latency of the entire clustering process. Furthermore, the confined subgraph traversal scheme improves the performance by an order of magnitude on undirected graphs, which is twice that of directed graphs.

Graph is an extremely versatile data structure in terms of its expressiveness and flexibility to model a range of real life phenomenon. Various networks like social networks, sensor networks and computer networks are represented and stored in the form of graphs. The analysis of these kind of graphs has an immense importance from quite a long time. It is performed from various aspects to get maximum out of such multifaceted information repository. When the analysis is targeted towards finding groups of vertices based on their similarity in a graph, clustering is the most conspicuous option. Previous graph clustering approaches either focus on the topological structure or attributes likeness, however, few recent methods constitutes both aspects simultaneously. Due to enormous computation requirements for similarity estimation, these methods are often suffered from scalability issues. In order to overcome this limitation, we introduce collaborative similarity measure (CSM) for intra-graph clustering. CSM is based on shortest path strategy, instead of all paths, to define structural and semantic relevance among vertices. First, we calculate the pair-wise similarity among vertices using CSM. Second, vertices are grouped together based on calculated similarity under k-Medoid framework. Empirical analysis, based on density, and entropy, proves the efficacy of CSM over existing measures. Moreover, CSM becomes a potential candidate for medium scaled graph analysis due to an order of magnitude less computations.

The all-inside arthroscopic modified Brostrom operation has been developed for lateral ankle instability. We compared the biomechanical parameters of the all-inside arthroscopic procedure to the open modified Brostrom operation. Eleven matched pairs of human cadaver specimens [average age 71.5 (range 58–98) years] were subject to the arthroscopic modified Brostrom operation using a suture anchor and the open modified Brostrom operation. The ligaments were loaded cyclically 20 times and then tested to failure. Torque to failure, degrees to failure, and stiffness were measured. A matched-pair analysis was performed. There was no significant difference in torque to failure between the open and arthroscopic modified Brostrom operation (19.9 ± 8.9 vs. 23.3 ± 12.1 Nm, n.s). The degrees to failure did not differ significantly between the open and arthroscopic modified Brostrom operations (46.8 ± 9.9° vs. 46.7 ± 7.6°, n.s). The working construct stiffness (or stiffness to failure) was no significant difference in the two groups (0.438 ± 0.21 vs. 0.487 ± 0.268 Nm/deg for the open and arthroscopic modified Brostrom operations, respectively, n.s). The all-inside arthroscopic modified Brostrom operation and the open modified Brostrom operation resulted in no significantly different torque to failure, degrees to failure, and working construct stiffness with no significant differences (n.s, n.s, and n.s, respectively). Our results indicate that the arthroscopic modified Brostrom operation is a reasonable alternative procedure for chronic ankle instability.

Pervasiveness of location acquisition technologies such as GPS, GSM, and Wi-Fi open the doors to use these technologies for ease and advancement of the society. One of the most emerging uses of these technologies is the low cost yet effective way of tracking of moving objects for activity monitoring. Daily lives of humans consist of enormous outdoor/trajectory activities like visiting different places for conducting routine task (e.g., Office, Restaurant, and Sports Club etc.). These activities put a significant effect in regulation of human life (i.e. health care and life care). By analyzing these activity traces and directing an effective routine of accomplishment of tasks can sufficiently improve its impact on human life. In this paper, we propose Daily Activity Monitor and LifeCare Provisioner (DALP) which is a GPS based outdoor activities analysis system for user monitoring and lifecare provisioning. To achieve real time and accurate outcome in tracking movement activities, we have proposed an approach of Personal tracking using static trajectory locations. DALP tracks the complete movement activity of a user and shares it with practitioners/instructors for analysis and updated recommendations. To verify and validate the working of DALP, a proof of concept prototype has been implemented that reflects its complete working.

Chronic disease may lead to life threatening health complications like heart disease, stroke, and diabetes that diminish the quality of life. CDSS (Clinical Decision Support System) helps physician in effective utilization of patient's clinical information at the time of diagnosis and medication. This paper points out the importance of social media and interaction integration in existing Smart CDSS for chronic diseases. The proposed system monitors health conditions, emotions and interests of patients from patients' tweets, trajectory and email analysis. We extract keywords, concepts and sentiments from patient's tweets data. Trajectory analysis identifies the focused activities after considering imperative location and semantic tags. Email analysis finds interesting patterns and communication trends from daily routine of patient. All these outputs are supplied to Smart CDSS into vMR (virtual Medical Record) format through social media adapter. This helps the health practitioners to understand the behavior and lifestyle of patients for better decision making about treatment. Consequently, patients can get continuous relevant recommendations from Smart CDSS based on their personalized profile. To verify and validate the working of proposed methodology, we have implemented a proof of concept prototype that reflects its complete working with potential outcomes.

Over the last few years, activity recognition in the smart home has become an active research area due to the wide range of human centric-applications. With the development of machine learning algorithms for activity classification, dataset is significantly important for algorithms testing and validation. Collection of real data is a challenging process due to involved budget, human resources, and annotation cost that’s why mostly researchers prefer to utilize existing datasets for evaluation purposes. However, openly available smart home datasets indicate variation in terms of performed activities, deployed sensors, and environment settings. Unfortunately, the analysis of existing datasets characteristic is a bottleneck for researchers while selecting datasets of their intent. In this paper, we develop a Framework for Smart Homes Dataset Analysis (FSHDA) to reflect their diverse dimensions in predefined format. It analyzes a list of data dimensions that covers the variations in time, activities, sensors, and inhabitants. For validation, we examine the effects of proposed data dimension on state-of-the-art activity recognition techniques. The results show that dataset dimensions highly affect the classifiers’ individual activity label assignments and their overall performances. The outcome of our study is helpful for upcoming researchers to develop a better understanding about the smart home datasets characteristics with classifier’s performance.

This paper deals with the problem of transforming eXtensible Markup Language (XML) data into the Resource Description Language (RDF) which can be understood by the computer. While it is not difficult to customize XML for arbitrary data, the effective transformation is not straightforward and the result may be not semantically richer than the source document since the redundancy data resulted from the duplicate elements in XML schema. To cope with this problem, we propose an approach to measure the similarity between these duplicates before giving the transforming strategy. The similarity measure is the combination of the children and ancestor factors, which describe the relationship of elements. The experimental results show that the proposed method gives the high degree of accuracy and produces better quality of RDF ontology.

A lot of research efforts have been made to model the diffusion process in social networks that varies from adoption of products in marketing strategies to disease and virus spread. Previously, a diffusion process is usually considered as a single-objective optimization problem, in which different heuristics or approximate algorithms are applied to optimize an objective of spreading single piece of information that captures the notion of diffusion. However, in real social networks individuals simultaneously receive several pieces of information during their communication. Single-objective solutions are inadequate for collective spread of several information pieces. Therefore, in this paper, we propose a Multi-Objective Diffusion Model (MODM) that allows the modeling of complex and nonlinear phenomena of multiple types of information exchange, and calculate the information worth of each individual from different aspects of information spread such as score, influence and diversity. We design evolutionary algorithm to achieve the multi-objectives in single diffusion process. Through extensive experiments on a real world data set, we have observed that MODM leads to a richer and more realistic class of diffusion model compared to a single objective. This signifies the correlation between the importance of each individual and his information processing capability. Our results indicate that some individuals in the network are naturally and significantly better connected in terms of receiving information irrespective of the starting position of the diffusion process.

Spectral clustering (SC) is currently one of the most popular clustering techniques because of its advantages over conventional approaches such as K-means and hierarchical clustering. However, SC requires the use of computing eigenvectors, making it time consuming. To overcome this limitation, Lin and Cohen proposed the power iteration clustering (PIC) technique (Lin and Cohen in Proceedings of the 27th International Conference on Machine Learning, pp. 655---662, 2010), which is a simple and fast version of SC. Instead of finding the eigenvectors, PIC finds only one pseudo-eigenvector, which is a linear combination of the eigenvectors in linear time. However, in certain critical situations, using only one pseudo-eigenvector is not enough for clustering because of the inter-class collision problem. In this paper, we propose a novel method based on the deflation technique to compute multiple orthogonal pseudo-eigenvectors (orthogonality is used to avoid redundancy). Our method is more accurate than PIC but has the same computational complexity. Experiments on synthetic and real datasets demonstrate the improvement of our approach.

Conventional procedures to ensure authorized data access by using access control policies are not suitable for cloud storage systems as these procedures can reveal valid access parameters to a cloud service provider. In this paper, we have proposed oblivious access control policy evaluation (O-ACE); a data sharing system, which obliviously evaluates access control policy on a cloud server and provisions access to the outsourced data. O-ACE reveals no useful information about the access control policy neither to the cloud service provider nor to the unauthorized users. Through the security analysis of O-ACE it has been observed that computational complexity to compromise privacy of the outsourced data is same as reverting asymmetric encryption without valid key pair. We have realized O-ACE for Google Cloud. Our evaluation results show the fact that O-ACE CPU utilization cost is 0.01–0.30 dollar per 1,000 requests.

Currently, a lot of recent electronic health records are based on XML documents. In order to integrate these heterogeneous XML medical documents efficiently, studies on finding structure and semantic similarity between XML Schemas have been exploited. The main problem is how to harvest the most appropriate relatedness to combine two schemas as a global XML Schema for reusing and referring purposes. In this paper, we propose the novel resemblance measure that concurrently considers both structural and semantic information of two specific healthcare XML Schemas. Specifically, we introduce new metrics to compute the datatype and cardinality constraint similarities, which improve the quality of the semantic assessment. On the basis of the similarity between each element pair, we put forward an algorithm to calculate the similarity between XML Schema trees. Experimental results lead to the conclusion that our methodology provides better similarity values than the others with regard to the accuracy of semantic and structure similarities.

Purpose This study assessed the clinical results and second-look arthroscopy after fibrin matrix-mixed gel-type autologous chondrocyte implantation to treat osteochondral lesions of the talus. Methods Chondrocytes were harvested from the cuboid surface of the calcaneus in 38 patients and cultured, and gel-type autologous chondrocyte implantation was performed with or without medial malleolar osteotomy. Preoperative American orthopedic foot and ankle society ankle-hind foot scores, visual analogue score, Hannover scoring system and subjective satisfaction were investigated, and the comparison of arthroscopic results (36/38, 94.7 %) and MRI investigation of chondral recovery was performed. Direct tenderness and relationship to the active daily life of the donor site was evaluated. Results The preoperative mean ankle–hind foot scores (71 ± 14) and Hannover scoring system (65 ± 10) had increased to 91 ± 12 and 93 ± 14, respectively, at 24-month follow-up (p

Activity recognition requires further research to enable a multitude of human-centric applications in the smart home environment. Currently, the major challenges in activity recognition include the domination of major activities over minor activities, their non-deterministic nature and the lack of availability of human-understandable output. In this paper, we introduce a novel Evolutionary Ensembles Model (EEM) that values both minor and major activities by processing each of them independently. It is based on a Genetic Algorithm (GA) to handle the non-deterministic nature of activities. Our evolutionary ensemble learner generates a human-understandable rule profile to ensure a certain level of confidence for performed activities. To evaluate the EEM, we performed experiments on three different real world datasets. Our experiments show significant improvement of 0.6 % to 0.28 % in the F-measures of recognized activities compared to existing counterparts. It is expected that EEM would be a practical solution for the activity recognition problem due to its understandable output and improved accuracy.

The aim of this work is to prospectively compare the effectiveness of iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL), T2-weighted fast spin-echo (FSE), and spoiled gradient-echo (SPGR) MR imaging to frequency selective fat suppression (FSFS) protocols for minimizing metallic artifacts in postoperative ankles with metallic hardware.The T2-weighted and SPGR imaging with IDEAL and FSFS were performed on 21 ankles of 21 patients with metallic hardware. Two musculoskeletal radiologists independently analyzed techniques for visualization of ankle ligaments and articular cartilage, uniformity of fat saturation, and relative size of the metallic artifacts. A paired t test was used for statistical comparisons of MR images between IDEAL and FSFS groups.IDEAL T2-weighted FSE and SPGR images enabled significantly improved visualization of articular cartilage (p0.05), the size of metallic artifact (p0.05), and the uniformity of fat saturation (p0.05). However, no significant improvement was found in the visibility of ligaments.IDEAL T2-weighted FSE and SPGR imaging effectively reduces the degree of tissue-obscuring artifacts produced by fixation hardware in ankle joints and improves image quality compared to FSFS T2-weighted FSE and SPGR imaging. However, visibility of ligaments was not improved using IDEAL imaging.

This paper addresses the issue of data governance in a cloud-based storage system. To achieve fine-grained access control over the outsourced data, we propose Self-Healing Attribute-based Privacy Aware Data Sharing in Cloud (SAPDS). The proposed system delegates the key distribution and management process to a cloud server without seeping out any confidential information. It facilitates data owner to restrain access of the user with whom data has been shared. User revocation is achieved by merely changing one attribute associated with the decryption policy, instead of modifying the entire access control policy. It enables authorized users to update their decryption keys followed by each user revocation, making it self-healing, without ever interacting with the data owner. Computation analysis of the proposed system shows that data owner can revoke n? users with the complexity of O(n?). Besides this, legitimate users can update their decryption keys with the complexity of O(1).

Compare the result of endoscopic versus open bursectomy in lateral malleolar bursitis.Prospective evaluation of 21 patients (22 ankles) undergoing either open or endoscopic excision of lateral malleolar bursitis. The median age was 64 (38-79) years old. The median postoperative follow-up was 15 (12-18) months.Those patients undergoing endoscopic excision showed a higher satisfaction rate (excellent 9, good 2) than open excision (excellent 4, good 3, fair 1). The wounds also healed earlier in the endoscopic group although the operation time was slightly longer. One patient in the endoscopic group had recurrence of symptoms but complications in the open group included one patient with skin necrosis, one patient with wound dehiscence, and two patients of with superficial peroneal nerve injury.Endoscopic resection of the lateral malleolar bursitis is a promising technique and shows favorable results compared to the open resection.Therapeutic studies-Investigating the result of treatment, Level II.

Acoustic source localization has many important applications particularly for military tracking foreign objects. Even though Wireless Sensor Networks (WSNs) have been developed, this localization problem remains a big challenge. A system for solving source localization must have the ability to deal with the problems of recorded convolved mixture signals while minimizing the high communication and computation cost. This paper introduces a distributed design for positioning multiple independent moving sources based on acoustic signals in which we focus on utilizing the relative information of magnitudes recorded at different sensors. The sensors perform preprocessing on the sensed data to capture the most important information before compressing and sending extracted data to the base. At the base, the data is uncompressed and the source locations are inferred via two clustering stages and an optimization method. Analysis and simulation results lead to the conclusion that our system provides good accuracy and needs neither much communication nor complex computation in a distributed manner. It works well when there exists high noise with Rayleigh multipath fading under Doppler effect and even when the number of independent sources is greater than the number of microphone sensors.

The topology of the trust network is important to optimize its usage in the trust-aware applications. However, since the users can join trust network ubiquitously, the structure of the highly dynamic trust network is still unknown. This paper contributes to verify that the trust network is the small-world network, and its small-world topology is independent of its dynamics. This is achieved by verifying the scale-freeness of five trust networks extracted from real online sites. Using the small-world nature of the trust network, we optimize the rating prediction mechanism of the conventional trust-aware recommender system. Experimental results clearly show that our proposed mechanism can achieve the maximum accuracy and coverage with the minimum computation complexity for the rating predictions.

This paper presents a new approach for identifying and eliminating mislabeled training instances for supervised learning algorithms. The novelty of this approach lies in the using of unlabeled instances to aid the detection of mislabeled training instances. This is in contrast with existing methods which rely upon only the labeled training instances. Our approach is straightforward and can be applied to many existing noise detection methods with only marginal modifications on them as required. To assess the benefit of our approach, we choose two popular noise detection methods: majority filtering (MF) and consensus filtering (CF). MFAUD/CFAUD is the new proposed variant of MF/CF which relies on our approach and denotes majority/consensus filtering with the aid of unlabeled data. Empirical study validates the superiority of our approach and shows that MFAUD and CFAUD can significantly improve the performances of MF and CF under different noise ratios and labeled ratios. In addition, the improvement is more remarkable when the noise ratio is greater.

Activity recognition is becoming an important research area, and finding its way to many application domains ranging from daily life services to industrial zones. Sensing hardware and learning algorithms are two important components in activity recognition. For sensing devices, we prefer to use accelerometers due to low cost and low power requirement. For learning algorithms, we propose a novel implementation of the semi-Markov Conditional Random Fields (semi-CRF) introduced by Sarawagi and Cohen. Our implementation not only outperforms the original method in terms of computation complexity (at least 10 times faster in our experiments) but also is able to capture the interdependency among labels, which was not possible in the previously proposed model. Our results indicate that the proposed approach works well even for complicated activities like eating and driving a car. The average precision and recall are 88.47% and 86.68%, respectively, which are higher than results obtained by using other methods such as Hidden Markov Model (HMM) or Topic Model (TM).

In this paper, we present a technique for estimating three-dimensional (3-D) human body posture from a set of sequential stereo images. We estimated the pixel displacements of stereo image pairs to reconstruct 3-D information. We modeled the human body with a set of ellipsoids connected by kinematic chains and parameterized with rotational angles at each body joint. To estimate human posture from the 3-D data, we developed a new algorithm based on expectation maximization (EM) with two-step iterations, assigning the 3-D data to different body parts and refining the kinematic parameters to fit the 3-D model to the data. The algorithm is iterated until it converges on the correct posture. Experimental results with synthetic and real data demonstrate that our method is capable of reconstructing 3-D human posture from stereo images. Our method is robust and generic; any useful information for locating the body parts can be integrated into our framework to improve the outcomes.

Wireless sensor networks (WSN) are susceptible to node capture and many network levels attacks. In order to provide protection against such threats, WSNs require lightweight and scalable key management schemes because the nodes are resource-constrained and high in number. Also, the effect of node compromise should be minimized and node capture should not hamper the normal working of a network. In this paper, we present an exclusion basis system-based key management scheme called MUQAMI+ for large-scale clustered sensor networks. We have distributed the responsibility of key management to multiple nodes within clusters, avoiding single points of failure and getting rid of costly inter-cluster communication. Our scheme is scalable and highly efficient in terms of re-keying and compromised node revocation.

Location awareness is the key capability of mobile computing applications. Despite high demand, indoor location technologies have not become truly ubiquitous mainly due to their requirements of costly infrastructure and dedicated hardware components. Received signal strength (RSS) based location systems are poised to realize economical ubiquity as well as sufficient accuracy for variety of applications. Nevertheless high resolution RSS based location awareness requires tedious sensor data collection and training of classifier which lengthens location system development life cycle. We present a rapid development approach based on online and incremental learning method which significantly reduces development time while providing competitive accuracy in comparison with other methods. ConSelFAM (Context-aware, Self-scaling Fuzzy ArtMap) extends the Fuzzy ArtMap neural network system. It enables on the fly expansion and reconstruction of location systems which is not possible in previous systems.

Ankle sprains are one of the most common sports injuries which are about 40% of total sports injuries and 20~40% of them would be progressed to chronic ankle instability (CAI) [1]. Rehabilitation and surgical therapy have been used to treat CAI, and the open modified Brostrm operation (MBO) is the gold standard surgical procedure. There are various evaluation methods in CAI treatments, such as the interview, visual analogue scale (VAS), and the measurement of range of motion (ROM) and the ankle torque. However, in reality, it is difficult to measure the maximum ROM and torque. Therefore, some studies measured ankle ROM and torque with cadaver specimens [2,3]. In this study, both open and arthroscopic MBO were performed on cadavers, and ankle torque and angle were measured during ankle inversion using the axial-torsion testing system. Ankle stiffness was calculated from measured data, and effects of both operations were compared quantitatively. For this study, matched pairs of fresh-frozen human cadaver lower leg specimens were obtained from seven males and four females (average age 71.5 (range 58–98) years). Each specimen consisted of the distal half of a leg. The soft tissues were removed from the calcaneus and distal tibiofibular part, except for the ankle joint and ligament. The anterior talofibular ligament (ATFL) and the calcaneofibular ligament (CFL) were transected. The specimens for the arthroscopic and open MBO were chosen from the left and right legs alternately. Then, each specimen was fixed in the specially-designed jig that was mounted on the axial-torsion fatigue testing system (Instron 8874, Norwood, MA, USA). The test consisted of a single ramp from 0° to 70° by inverting the ankle at 5°/s, while measuring the angular position and resultant torque. There was no statistical difference in torque to failure between open and arthroscopic MBO. The maximum torque was 16.3±7.2 N•m at 45° for open MBO and 19.9±10.8 N•m at 40° for arthroscopic MBO (Figure ​(Figure1).1). Ankle stiffness increased faster in arthroscopic MBO (0.31±0.22 N•m/deg) than in open MBO (0.18±0.15 N•m/deg) in initial inversion range (