Broadcasting is an important operation and been widely used in wireless sensor networks （WSNs）. These networks are power constrained as nodes operate with limited battery power. Wireless sensor networks are spatial ...Broadcasting is an important operation and been widely used in wireless sensor networks （WSNs）. These networks are power constrained as nodes operate with limited battery power. Wireless sensor networks are spatial graphs that have much more clustered and much high path-length characteristics. After considering energy- efficient broadcasting in such networks, by combining the small-world characteristic of WSNs and the properties of ant algorithm to quickly identify an optimal path, small-world power-aware broadcast algorithm is introduced and evaluated. Given different densities of network, simulation results show that our algorithm significantly improves life of networks and also reduces communication distances and power consumption.展开更多
In order to solve the ambiguity problems in the semantic context （structure, granularity or scale） emerging in the process of ontology integration application, this paper analyzes the essential characters of context...In order to solve the ambiguity problems in the semantic context （structure, granularity or scale） emerging in the process of ontology integration application, this paper analyzes the essential characters of context structure, proposes a novel semantic context generating algorithm, which is implemented over VO-Editor（visual ontology editor）, from the satisfiability-based point of view, and proves that the context entity generated by this algorithm is smallest in scale and unique. It offers a feasible means for developers to handle context problems for ontology integration application.展开更多
In order to improve the performance of classifiers in subjective domains, this paper defines a metric to measure the quality of the subjectively labelled training data (QoSTD) by means of K-means clustering. Then, the...In order to improve the performance of classifiers in subjective domains, this paper defines a metric to measure the quality of the subjectively labelled training data (QoSTD) by means of K-means clustering. Then, the QoSTD is used as a weight of the predicted class scores to adjust the likelihoods of instances. Moreover, two measurements are defined to assess the performance of the classifiers trained by the subjective labelled data. The binary classifiers of Traditional Chinese Medicine (TCM) Zhengs are trained and retrained by the real-world data set, utilizing the support vector machine (SVM) and the discrimination analysis (DA) models, so as to verify the effectiveness of the proposed method. The experimental results show that the consistency of likelihoods of instances with the corresponding observations is increased notable for the classes, especially in the cases with the relatively low QoSTD training data set. The experimental results also indicate the solution how to eliminate the miss-labelled instances from the training data set to re-train the classifiers in the subjective domains.展开更多
In order to evaluate the structural complexity of class diagrams systematically and deeply, a new guiding framework of structural complexity is presented. An index system of structural complexity for class diagrams is...In order to evaluate the structural complexity of class diagrams systematically and deeply, a new guiding framework of structural complexity is presented. An index system of structural complexity for class diagrams is given. This article discusses the formal description of class diagrams, and presents the method of formally structural complexity metrics for class diagrams from associations, dependencies, aggregations, generalizations and so on. An applicable example proves the feasibility of the presented method.展开更多
1 Introduction With the rapid development of mobile networks,locationbased services has become popular in the daily lives of people.The service providers can recommend the profitable services to persons through mining...1 Introduction With the rapid development of mobile networks,locationbased services has become popular in the daily lives of people.The service providers can recommend the profitable services to persons through mining the frequent interests or places of persons.However,one aspect is that the historical data on Internet can easily cause the leakage of user-relationship privacy,another aspect is that the historical interests of person are always bound to time.Therefore,this paper devotes to study a privacy protection method on time-constrained point of interests(PoIs)based on the group relationships of users.展开更多
By considering energy-efficient anycast routing in wireless sensor network （WSN）, and combining small world characteristics of WSN with the properties of the ant algorithm, a power-aware anycast routing algorithm （...By considering energy-efficient anycast routing in wireless sensor network （WSN）, and combining small world characteristics of WSN with the properties of the ant algorithm, a power-aware anycast routing algorithm （SWPAR） with multi-sink nodes is pro- posed and evaluated. By SWPAR, the optimal sink node is found and the problem of routing path is effectively solved. Simulation results show that compared with the sink-based anycast routing protocol （SARP） and the hierarchy-based anyeast routing protocol （HARP）, the proposed algorithm improves network lifetime and reduces power consumption.展开更多
Diet plays an important role in people’s daily life with its strong correlation to health and chronic diseases. Meanwhile, deep based food computing emerges to provide lots of works which including food recognition, ...Diet plays an important role in people’s daily life with its strong correlation to health and chronic diseases. Meanwhile, deep based food computing emerges to provide lots of works which including food recognition, food retrieval, and food recommendation, and so on. This work focuses on the food recognition, specially, the ingredients identification from food images. The paper proposes two types of ways for ingredient identification. Type1 method involves the combination of salient ingredients classifier with salient ingredient identifiers. Type 2 method introduces the segment-based classifier. Furthermore, this work chooses 35 kinds of ingredients in the daily life as identification categories, and constructs three kinds of novel datasets for establishing the ingredient identification models. All of the classifiers and identifiers are trained on Resnet50 by transfer learning. Many experiments are conducted to analyze the effectiveness of proposed methods. As the results, Salient ingredients classifier predict one ingredient and achieves 91.97% on test set of salient ingredients dataset and 82.48% on test dish image dataset. Salient ingredients identifiers predict remained ingredients and achieve mean accuracy of 85.96% on test dish image dataset. Furthermore, Segment-based classifier achieves 94.81% on test set of segment-based ingredients dataset.展开更多
How to organize crossing social network resources on a higher level of integration and address them to users＇ desktops is an important difficult problem. Especially, there is a lack of efficient approaches to softwar...How to organize crossing social network resources on a higher level of integration and address them to users＇ desktops is an important difficult problem. Especially, there is a lack of efficient approaches to software architecture to build reusable system over the crossing social network, From the viewpoint of temporal logic XYZ/E, this paper proposes a kind of Architecture Descrip- tion Language about the Crossing Social Network system （CSN_ADL）, which can be used to depict the main key processes over the cross-social network system, and formally defines some key concepts, such as relation component, corelation component, override corelation connector, interaction connector, corelation network-oriented architecture, as well as system correctness, system activity, and system safety. Furthermore, some properties of correctness, activity, and safety under the flame CSN_ADL is discussed and depicted formally, which provides a formally theo- retical instruction for architecture reuses.展开更多
This paper dealt with composite scheduling problems which combine manufacturing scheduling problems and/or transportation routing problems.Two scheduling models were formulated as the elements of the composite schedul...This paper dealt with composite scheduling problems which combine manufacturing scheduling problems and/or transportation routing problems.Two scheduling models were formulated as the elements of the composite scheduling model,and the composite model was formulated composing these models with indispensable additional constraints.A hybrid genetic algorithm was developed to solve the composite scheduling problems.An improved representation based on random keys was developed to search permutation space.A genetic algorithm based dynamic programming approach was applied to select resource.The proposed technique and a previous technique are compared by three types of problems.All results indicate that the proposed technique is superior to the previous one.展开更多
The problem of subgraph matching is one fundamental issue in graph search,which is NP-Complete problem.Recently,subgraph matching has become a popular research topic in the field of knowledge graph analysis,which has ...The problem of subgraph matching is one fundamental issue in graph search,which is NP-Complete problem.Recently,subgraph matching has become a popular research topic in the field of knowledge graph analysis,which has a wide range of applications including question answering and semantic search.In this paper,we study the problem of subgraph matching on knowledge graph.Specifically,given a query graph q and a data graph G,the problem of subgraph matching is to conduct all possible subgraph isomorphic mappings of q on G.Knowledge graph is formed as a directed labeled multi-graph having multiple edges between a pair of vertices and it has more dense semantic and structural features than general graph.To accelerate subgraph matching on knowledge graph,we propose a novel subgraph matching algorithm based on subgraph index for knowledge graph,called as FGqT-Match.The subgraph matching algorithm consists of two key designs.One design is a subgraph index of matching-driven flow graph(FGqT),which reduces redundant calculations in advance.Another design is a multi-label weight matrix,which evaluates a near-optimal matching tree for minimizing the intermediate candidates.With the aid of these two key designs,all subgraph isomorphic mappings are quickly conducted only by traversing FGqj.Extensive empirical studies on real and synthetic graphs demonstrate that our techniques outperform the state-of-the-art algorithms.展开更多
文摘Broadcasting is an important operation and been widely used in wireless sensor networks （WSNs）. These networks are power constrained as nodes operate with limited battery power. Wireless sensor networks are spatial graphs that have much more clustered and much high path-length characteristics. After considering energy- efficient broadcasting in such networks, by combining the small-world characteristic of WSNs and the properties of ant algorithm to quickly identify an optimal path, small-world power-aware broadcast algorithm is introduced and evaluated. Given different densities of network, simulation results show that our algorithm significantly improves life of networks and also reduces communication distances and power consumption.
基金the National Natural Science Foundation of China ( 90604005)
文摘In order to solve the ambiguity problems in the semantic context （structure, granularity or scale） emerging in the process of ontology integration application, this paper analyzes the essential characters of context structure, proposes a novel semantic context generating algorithm, which is implemented over VO-Editor（visual ontology editor）, from the satisfiability-based point of view, and proves that the context entity generated by this algorithm is smallest in scale and unique. It offers a feasible means for developers to handle context problems for ontology integration application.
文摘In order to improve the performance of classifiers in subjective domains, this paper defines a metric to measure the quality of the subjectively labelled training data (QoSTD) by means of K-means clustering. Then, the QoSTD is used as a weight of the predicted class scores to adjust the likelihoods of instances. Moreover, two measurements are defined to assess the performance of the classifiers trained by the subjective labelled data. The binary classifiers of Traditional Chinese Medicine (TCM) Zhengs are trained and retrained by the real-world data set, utilizing the support vector machine (SVM) and the discrimination analysis (DA) models, so as to verify the effectiveness of the proposed method. The experimental results show that the consistency of likelihoods of instances with the corresponding observations is increased notable for the classes, especially in the cases with the relatively low QoSTD training data set. The experimental results also indicate the solution how to eliminate the miss-labelled instances from the training data set to re-train the classifiers in the subjective domains.
基金Science and Technology Department Term of Education of Heilongjiang Province(Grant No.11511127)
文摘In order to evaluate the structural complexity of class diagrams systematically and deeply, a new guiding framework of structural complexity is presented. An index system of structural complexity for class diagrams is given. This article discusses the formal description of class diagrams, and presents the method of formally structural complexity metrics for class diagrams from associations, dependencies, aggregations, generalizations and so on. An applicable example proves the feasibility of the presented method.
基金This work was supported by the National Natural Science Foundation of China(Grant Nos.61976032 and 62002039)the General Scientific Research Project of Liaoning(No.LJKZ0063).
文摘1 Introduction With the rapid development of mobile networks,locationbased services has become popular in the daily lives of people.The service providers can recommend the profitable services to persons through mining the frequent interests or places of persons.However,one aspect is that the historical data on Internet can easily cause the leakage of user-relationship privacy,another aspect is that the historical interests of person are always bound to time.Therefore,this paper devotes to study a privacy protection method on time-constrained point of interests(PoIs)based on the group relationships of users.
基金The Grand Fundamental Advanced Research of Chinese National Defense (No.S0500A001)
文摘By considering energy-efficient anycast routing in wireless sensor network （WSN）, and combining small world characteristics of WSN with the properties of the ant algorithm, a power-aware anycast routing algorithm （SWPAR） with multi-sink nodes is pro- posed and evaluated. By SWPAR, the optimal sink node is found and the problem of routing path is effectively solved. Simulation results show that compared with the sink-based anycast routing protocol （SARP） and the hierarchy-based anyeast routing protocol （HARP）, the proposed algorithm improves network lifetime and reduces power consumption.
文摘Diet plays an important role in people’s daily life with its strong correlation to health and chronic diseases. Meanwhile, deep based food computing emerges to provide lots of works which including food recognition, food retrieval, and food recommendation, and so on. This work focuses on the food recognition, specially, the ingredients identification from food images. The paper proposes two types of ways for ingredient identification. Type1 method involves the combination of salient ingredients classifier with salient ingredient identifiers. Type 2 method introduces the segment-based classifier. Furthermore, this work chooses 35 kinds of ingredients in the daily life as identification categories, and constructs three kinds of novel datasets for establishing the ingredient identification models. All of the classifiers and identifiers are trained on Resnet50 by transfer learning. Many experiments are conducted to analyze the effectiveness of proposed methods. As the results, Salient ingredients classifier predict one ingredient and achieves 91.97% on test set of salient ingredients dataset and 82.48% on test dish image dataset. Salient ingredients identifiers predict remained ingredients and achieve mean accuracy of 85.96% on test dish image dataset. Furthermore, Segment-based classifier achieves 94.81% on test set of segment-based ingredients dataset.
基金Supported by the Fujian Province Science Research Foundation Grant (2009J01272)the Research Fund (type A) (JA09038) from the Education Department of Fujian Provincethe Humanities and Social Science Research Projects of the Ministry of Education (11YJA860028)
文摘How to organize crossing social network resources on a higher level of integration and address them to users＇ desktops is an important difficult problem. Especially, there is a lack of efficient approaches to software architecture to build reusable system over the crossing social network, From the viewpoint of temporal logic XYZ/E, this paper proposes a kind of Architecture Descrip- tion Language about the Crossing Social Network system （CSN_ADL）, which can be used to depict the main key processes over the cross-social network system, and formally defines some key concepts, such as relation component, corelation component, override corelation connector, interaction connector, corelation network-oriented architecture, as well as system correctness, system activity, and system safety. Furthermore, some properties of correctness, activity, and safety under the flame CSN_ADL is discussed and depicted formally, which provides a formally theo- retical instruction for architecture reuses.
基金Project supported by the Grant-in-Aid for Young Scientists (B) from the Ministry of Education,Culture,Sports,Science and Technology,Japan
文摘This paper dealt with composite scheduling problems which combine manufacturing scheduling problems and/or transportation routing problems.Two scheduling models were formulated as the elements of the composite scheduling model,and the composite model was formulated composing these models with indispensable additional constraints.A hybrid genetic algorithm was developed to solve the composite scheduling problems.An improved representation based on random keys was developed to search permutation space.A genetic algorithm based dynamic programming approach was applied to select resource.The proposed technique and a previous technique are compared by three types of problems.All results indicate that the proposed technique is superior to the previous one.
基金the National Natural Science Foundation of China(Grant Nos.61976032,62002039).
文摘The problem of subgraph matching is one fundamental issue in graph search,which is NP-Complete problem.Recently,subgraph matching has become a popular research topic in the field of knowledge graph analysis,which has a wide range of applications including question answering and semantic search.In this paper,we study the problem of subgraph matching on knowledge graph.Specifically,given a query graph q and a data graph G,the problem of subgraph matching is to conduct all possible subgraph isomorphic mappings of q on G.Knowledge graph is formed as a directed labeled multi-graph having multiple edges between a pair of vertices and it has more dense semantic and structural features than general graph.To accelerate subgraph matching on knowledge graph,we propose a novel subgraph matching algorithm based on subgraph index for knowledge graph,called as FGqT-Match.The subgraph matching algorithm consists of two key designs.One design is a subgraph index of matching-driven flow graph(FGqT),which reduces redundant calculations in advance.Another design is a multi-label weight matrix,which evaluates a near-optimal matching tree for minimizing the intermediate candidates.With the aid of these two key designs,all subgraph isomorphic mappings are quickly conducted only by traversing FGqj.Extensive empirical studies on real and synthetic graphs demonstrate that our techniques outperform the state-of-the-art algorithms.
