Search

Generally, QA systems suffer from the structural difference where a question is composed of unstructured data, while its answer is made up of structured data in a Knowledge Graph (KG). To bridge this gap, most approaches use lexicons to cover data that are represented differently. However, the existing lexicons merely deal with representations for entity and relation mentions rather than consulting the comprehensive meaning of the question. To resolve this, we design a novel predicate constraints lexicon which restricts subject and object types for a predicate. It facilitates a comprehensive validation of a subject, predicate and object simultaneously. In this paper, we propose Predicate Constraints based Question Answering (PCQA). Our method prunes inappropriate entity/relation matchings to reduce search space, thus leading to an improvement of accuracy. Unlike the existing QA systems, we do not use any templates but generates query graphs to cover diverse types of questions. In query graph generation, we put more focus on matching relations rather than linking entities. This is well-suited to the use of predicate constraints. Our experimental results prove the validity of our approach and demonstrate a reasonable performance compared to other methods which target WebQuestions and Free917 benchmarks.

One of the noticeable characteristics of the Internet-of-Things (IoT) devices is that they are resource-constrained, which makes them incompatible with the standard Internet protocols, e.g., HTTP. Nevertheless, IoT offers the Constrained Application Protocol (CoAP) as an alternative protocol for IoT devices where it is considered to be lightweight in regard to power consumption, network traffic, and so on. The main challenges associated with CoAP, however, still remain unsolved where most of the state-of-the-art approaches group servers manually and return the exact matches only. To address these issues, we propose a novel Semantically Enhanced CoAP Gateway (SECoG) for complex service mashups. SECoG enables users to create a simple or complex CoAP service mashup via a single HTTP request. Furthermore, a prototype implementation of SECoG was deployed on a real-world testbed to evaluate the proposed approach in terms of the reliability, execution time, network traffic, and accuracy.

The sharp growth of distributed energy-related resources requires an efficient energy management for future grids. The traditional power grid that highly depends on information model standards collects energy data depending on them and creates energy services. Nowadays, decentralized grids utilize information schema generated by reusing standards as well as existing schemas. The schema helps implement smart energy services in the future grids. To meet such requirements, domain experts proposed upper-level schemas that manage a wide range of energy-related knowledge resources. However, their schemas could not conduct an effective reuse of energy-related knowledge resources due to their unsuitable schema development methodologies. Moreover, there is a lack of vocabularies that satisfy critical requirements for decentralized grids. To cope with these problems, we propose an energy knowledge graph (EKG) as an upper schema for the integration of knowledge resources in energy systems. First, we utilize the existing methodology that offers guidelines for reusing existing knowledge resources. Second, EKG supports concepts of energy trading and communities to satisfy the requirements of decentralized grids. Third, EKG presents compliant concepts that are compatible with existing schemas. Fourth, we modeled the use cases using the EKG and evaluated them according to the scenario specification of energy services. Last, to demonstrate the benefits of the EKG, we implemented key components such as semantic mashup and complex event processing.

Recently, the emerging Social Internet of Things (SIoT) has opened up a myriad of research opportunities. One of the most fundamental research challenges posed by SIoT and its objective is to allow smart objects to create and maintain their own social networks. In this paper, we are interested in constructing these social networks, which are built upon social relationships between smart objects, and quantitatively estimating the social strength values of the relations. In particular, we propose a model named social strength prediction model, which infers social connections among smart objects and predicts the strength of the connections using the co-usage data of the objects. The proposed model is divided into two major components: 1) entropy-based and 2) distance-based social strength computations. The two components capture different properties of co-usages of objects, namely, the diversity and spatiotemporal features, which are all essential factors that contribute to the values of social strength. In order to test the feasibility of the proposed model, we conducted extensive sets of experiments with real-world datasets, which include the history data of various object usages. Based on the results, we show that the proposed model is in fact capable of inferring social connections and predicting their corresponding social strength with relatively high precision and recall.

Context-aware recommender systems (CARS) exploit multiple contexts to improve user experience in embracing new information and services. Tensor factorization (TF), a type of latent factor model, has achieved remarkable performance in CARS. TF learns latent representations of contexts by decomposing an observed rating tensor and combines the latent representations as a vector form to represent contextual influence on users and items. However, due to the limitation of the contextual expression power, they have difficulties in effectively capturing complex correlations among multiple contexts, and also the meaning of each context is diluted. To address the issue, we propose a reliable TF-based recommender system based on a proposed context tensor (CT-CARS), which incorporates a variety of correlations among contexts. CT-CARS contains a novel recommendation rating function and a learning algorithm. Specifically, the proposed context tensor elaborately captures the influences of both individual contexts and context combinations. Moreover, we introduce a novel parameter initialization based on past-learned results to improve the reliability of recommendations. The overall time complexity of our parameter learning algorithm grows linearly as dataset size increases. Experiments on six real-world datasets including two large-scaled datasets show that CT-CARS outperforms the existing state-of-the-art models in terms of both accuracy and reliability.

To satisfy a user’s complex requirements, Resource Description Framework (RDF) Stream Processing (RSP) systems envision the fusion of remote RDF data with semantic streams, using common data models to query semantic streams continuously. While streaming data are changing at a high rate and are pushed into RSP systems, the remote RDF data are retrieved from different remote sources. With the growth of SPARQL endpoints that provide access to remote RDF data, RSP systems can easily integrate the remote data with streams. Such integration provides new opportunities for mixing static (or quasi-static) data with streams on a large scale. However, the current RSP systems do not offer any optimisation for the integration. In this article, we present an adaptive plan-based approach to efficiently integrate sematic streams with the static data from a remote source. We create a query execution plan based on temporal constraints among constituent services for the timely acquisition of remote data. To predict the change of remote sources in real time, we propose an adaptive process of detecting a source update, forecasting the update in the future, deciding a new plan to obtain remote data and reacting to a new plan. We extend a SPARQL query with operators for describing the multiple strategies of the proposed adaptive process. Experimental results show that our approach is more efficient than the conventional RSP systems in distributed settings.

Recent advances in the Internet of Things (IoT) have led to the rise of a new paradigm: Social Internet of Things (SIoT). However, the new paradigm, as inspired by the idea that smart objects will soon have a certain degree of social consciousness, is still in its infant state for several reasons. Most of the related works are far from embracing the social aspects of smart objects and the dynamicity of inter-object social relations. Furthermore, there is yet to be a coherent structure for organising and managing IoT objects that elicit social-like features. To fully understand how and to what extent these objects mimic the behaviours of humans, we first model SIoT by scrutinising the distinct characteristics and structural facets of human-centric social networks. To elaborate, we describe the process of profiling the IoT objects that become social and classify various inter-object social relationships. Afterwards, a novel discovery mechanism, which utilises our hypergraph-based overlay network model, is proposed. To test the feasibility of the proposed approach, we have performed several experiments on our smart home automation demo box built with various sensors and actuators.

Information Systems (ISs) have become one of the crucial tools for various organizations in managing and coordinating business processes. Now we are entering the era of the Internet of Things (IoT). IoT is a paradigm in which real-world physical things can be connected to the Internet and provide services through the computing devices attached. The IoT infrastructure is starting to be integrated with ISs thereby diminishing the boundaries between the physical world and the business IT systems. With the development of IoT technologies, the number of connected things and their available physical services are increasing rapidly. Thus, selecting an appropriate service that satisfies a user’s requirements from such services becomes a time-consuming challenge. To address this issue, we propose a Physical Service Model (PSM) as a common conceptual model to describe heterogeneous IoT physical services. PSM contains three core concepts (device, resource, and service) and specifies their relationships. Based on the proposed PSM, we define three types of Quality of Service (QoS) attributes and rate candidate services according to user requirements. To dynamically rate QoS values and select an appropriate physical service, we propose a Physical Service Selection (PSS) method that takes a user preference and an absolute dominance relationship among physical services into account. Finally, experiments are conducted to evaluate the performance of the proposed method.

Objectives To enhance myogenic differentiation in pulp cells isolated from extracted premolars by epigenetic modification using a DNA demethylation agent, 5-aza-2'-deoxycytidine (5-Aza), and to evaluate the potent stimulatory effect of 5-Aza-treated pulp cell injection for craniofacial muscle regeneration in vivo. Setting and sample population Pulp cells were isolated from premolars extracted for orthodontic purposes from four adults (age range, 18-22.1 years). Material and methods Levels of myogenic differentiation and functional contraction response in vitro were compared between pulp cells with or without pre-treatment of 5-Aza. Changes in muscle regeneration in response to green fluorescent protein (GFP)-labelled myogenic pulp cell injection in vivo were evaluated using a cardiotoxin (CTX)-induced muscle injury model of the gastrocnemius as well as the masseter muscle in mice. Results Pre-treatment of 5-Aza in pulp cells stimulated myotube formation, myogenic differentiation in terms of desmin and myogenin expression, and the level of collagen gel contraction. The local injection of 5-Aza pre-treated myogenic pulp cells was engrafted into the host tissue and indicated signs of enhanced muscle regeneration in both the gastrocnemius and the masseter muscles. Conclusion The epigenetic modification of pulp cells from extracted premolars and the local injection of myogenic pulp cells may stimulate craniofacial muscles regeneration in vivo.

Fog computing is a promising paradigm in terms of extending cloud computing to an edge network. In a broad sense, fog computing in Internet-of-Vehicles(IoV) provides low-latency services since fog nodes are closely located with moving cars and are locally distributed. In this paper, we propose a fog computing architecture based on a publish/subscribe model. After that, we describe a traffic congestion control scenario using a smart traffic light system which operates on top of the proposed architecture. Furthermore, we propose an upper-level domain ontology in order to enhance the expressivity of knowledge and describe a variety of semantic properties which interlink spatial information in IoV. Finally, we present an active rule where supports the exchange of event-driven messages between publishing and subscribing fog nodes.

RDF stream processing (RSP), which aims to query data streams and linked datasets using common data model and query languages extended from RDF and SPARQL, is gaining popularity. However, most of the existing RSP engines do not provide any optimisation techniques for shared join operators among query plans from concurrent queries. Many number of shared join operators can incur the waste of a lot of CPU resources like a processing memory. Moreover, queries on shared operators cause a slow response time because they must be re-evaluated without reusing intermediate results. To solve these problems, we propose an efficient method of optimising query plans on multiple queries. First, the proposed method evicts some data that get notified from the streams in order to maintain an efficient memory usage. Second, the proposed method optimises query plans to maximise the reuse of shared join results. Experimental results show that the proposed method has significant improvements in terms of memory consumption and latency, compared to the state-the-of-art methods.

As the Internet of Things (IoT) is gaining tremendous popularity from industries to researchers for its potential to provide business opportunities and academic contributions, several research challenges have risen to the surface. With billions of IoT objects being produced, connected and interacting, one of the challenges claim that there is an imperative need for ways to organize and manage the objects effectively. This paper, therefore, proposes a network-based multidimensional structure to model the IoT for efficient management and discovery of IoT objects. In addition, we address various research issues that must be considered in designing the proposed model. To test the feasibility of the model, we present numerous experiments performed on our smart home automation demo box, along with their justifications.

The Internet of Things (IoT) is a paradigm where real-world entities can be connected to the Internet and provide services with the help of attached devices. With the development of IoT technologies, the number of devices deployed around the world as well as their services are increasing rapidly. Thus, selecting an appropriate service which satisfies a user's requirements from such many services becomes a time consuming challenge. Therefore, to address this issue we propose a Physical Service Model (PSM) to describe heterogeneous IoT physical services. PSM contains three core concepts (Device, Resource and Service) and specifies their relationships. Based on PSM, we define four quality of service (QoS) attributes and rate candidate services according to user requirements. In order to dynamically aggregate individual QoS ratings and select physical services, we propose a Physical Service Selection (PSS) method which takes a user preference and an absolute dominance relationship among physical services into account. Finally, experiments are conducted to evaluate the performance of the proposed method.

Most of the mobile platforms provide a keyword based full text search (FTS) for users to find what they want. However, FTS has difficulties in dealing with the cases where a user cannot remember the exact keywords about target data or the number of search results is too many. To overcome these limitations of FTS, we propose a semantically enhanced method of searching for data on mobile devices along with mobile ontology. Experimental results of the proposed method show that our method provides accurate search results and is suitable for a mobile environment.

To efficiently explore heterogeneous objects on the Internet and uniquely identify them, we propose a scalable physical-object naming system that reuses the existing ontologies and assigns URL-based semantic identifiers. We implement a prototype of the proposed system and evaluate its performance using synthetically generated data based on the W3C Semantic Sensor Network ontology.

To enhance the integration of different IoT application domains, we propose a two-layered IoT information model that consists of domain-independent and domain-specific models. In addition, for the meaningful composition of things and their services, a semantic URI as a unique identifier is allocated to a thing. Finally, a modeling approach of constructing physical mashups is proposed for the event-driven composition of IoT services. In order to show the efficiency of the proposed event-driven physical mashup approach, an energy management scenario in u-campus is demonstrated as a case study.

To enhance the incorrectness of keyword based search, we propose an efficient semantic search method based on a lightweight mobile ontology designed for smart mobile devices. In addition, we implement a prototype of semantic search engine working on Android smartphones and our prototype engine provides better user experience compared with keyword based search.

A mobile mash up should be able to compose the component services in accordance with a user's context. We propose a context-aware mash up method for smart mobile devices. For this purpose, a novel user request model is presented and the conventional service description model is extended. We propose a context ontology which can efficiently manage contextual information in a mobile environment. Furthermore, the proposed method undergoes three computations -- functional, I/O connectable and contextual matching. In particular, the contextual matching technique utilizes two types of contextual information that are proposed in this paper, implicit and explicit contexts. In order to evaluate the accuracy of the proposed method, a test registry based on real world Web APIs was built. Experimental results show an overall improvement of 26.7% in precision compared with the case where contextual information was not utilized.

There are difficulties in applying role-based dynamic delegation to Web services. This is because XACML, which is a standard language for describing the access policies of Web services, does not support the dynamic delegation between roles in different domains. We propose role attributes to specify the characteristics of roles in different domains to support role-based dynamic delegation in a Web services system, which services in different administrative domains work together. The proposed method enables the dynamic delegation in a service-oriented system based on the proposed role attributes. Experimental results show that the proposed method performs better than previous methods, resulting in enhancing an authorization rate about 32 percent on average.

Social networks facilitate information sharing and communication among people who have common interests. In the field of services computing, many researches have tried to exploit social networks to compute the reputations of web services. However, the reputation scores computed in most of these studies are still shown to be unreliable and lack trustworthiness. Therefore, this paper proposes an Integrated Social Service Network Model IS2NM which combines social and service network models along with the interactions among different types of service participants. Moreover, we present a sophisticated method of computing service reputations, which considers various types of participants such as service requesters, raters and providers participating in the service ecosystem. Based on the model, the proposed method calculates the credibility and expertise of the raters and providers of a service and incorporates them to the computation of service reputation scores. Experimental results with real-world web services show that the proposed method computes service reputations more elaborately and precisely.