Textual Knowledge Flow (TKF) provides an effective technique and theoretical support for intelligent browsing. In order to realize the intelligent browsing of topics in Web or Library, TKF based intelligent browsing o...
详细信息
Textual Knowledge Flow (TKF) provides an effective technique and theoretical support for intelligent browsing. In order to realize the intelligent browsing of topics in Web or Library, TKF based intelligent browsing of topics is proposed. Firstly, topics are represented by Element Fuzzy Cognitive Maps (E-FCMs); then semantic values of concepts and semantic influences of relations are calculated based on their frequencies and weights in E-FCMs library. Thirdly, semantic similarity degrees between topics are calculated for building the semantic Link Network (SLN). Fourthly, with the help of SLN and user's demand, TKF between similar topics is activated as browsing path of topics to guide user's browsing behavior. Experimental results show that the browsing path of topics is easy to be built by the proposed method. TKF based topic browsing has a brilliant perspective in the applications of intelligent browsing, knowledge grid and semantic Web.
P2P networks have become a popular way to share large volumes of data due to its open and anonymous nature. P2P network model is designed with the targets of decentralization, well-scaling, fault-tolerance and low cos...
详细信息
Traditional text mining techniques have weak ability to provide associated relations with rich semantics that is a foundation of the intelligent browsing of topics, discovery of semantic community and precise personal...
详细信息
Traditional text mining techniques have weak ability to provide associated relations with rich semantics that is a foundation of the intelligent browsing of topics, discovery of semantic community and precise personalized recommendation in current Web and knowledge grid, etc. In this paper we propose an algorithm to generate and calculate the associated relations and their strengths between documents within a domain. Each document is represented by a bag of words and their weights. We first build domain knowledge background based on the association rules at keyword level, and then we apply those association rules to generate and calculate the documents' semantic relations and their strengths at document level, which effectively shorten the semantic gap from keyword semantics to document semantics. Experimental results show that our proposed method is feasible and able to discover interesting facts within a domain.
Intelligent browsing of topics is one of the key issues of the Web and the knowledge grid, which provides on-demand services to support knowledge discovery and innovation. In this paper, Background of topics composed ...
详细信息
ISBN:
(纸本)9781424418657
Intelligent browsing of topics is one of the key issues of the Web and the knowledge grid, which provides on-demand services to support knowledge discovery and innovation. In this paper, Background of topics composed by Associated Relations (BAR) between keywords belonging to one domain is developed to extend the topicspsila associated relations based on the element fuzzy cognitive maps (E-FCMs), which helps us discover the associated topics that do not have direct associated relation at the keyword level. The matrix of element concepts (MC) is presented to store element concepts of topics. Through MC we can obtain the causal concepts of each topic in one domain; logic operation ldquoANDrdquo is given to get associated topics. semantic Link Network (SLN) is automatically generated by the associated topics, which can realize the intelligent browsing of topics in Web or knowledge grid. Experimental results show that the browsing path of topics is easier to be built by the proposed algorithm and the associated topic can be recommended automatically. Meanwhile, an application of the associated topics in multimedia provides a good prospect on the cross media intelligent browsing.
Communications among web services are asynchronous. Asynchronous models of service compositions face the problem that the performance of verification is bring down with states explosion. We describe an approach, which...
详细信息
In the Web personalization of Web service and construction of adaptive Web site, how to represent uses profile is one of the key issues. Aiming at solving the existing problems in representation of user profile includ...
详细信息
ISBN:
(纸本)9780769533162
In the Web personalization of Web service and construction of adaptive Web site, how to represent uses profile is one of the key issues. Aiming at solving the existing problems in representation of user profile includes incapability of effectively representing userpsilas recent interest, lack in dynamically effectively updating. This paper presents a new method of representation of user profile. By computing the interest degree with association degree between topics, updating of interest degree is preformed. By introducing memory model, it can predict which Web pages are still concerned by the user. By classifying the operations on the presented user profile, the construction and evolution can be done. In case study a userpsilas browsing activities is traced and corresponding user profile is built and evolves, thus to prove the feasibility of this method.
P2P networks have become a popular way to share large volumes of data due to its open and anonymous nature. P2P network model is designed with the targets of decentralization, well-scaling, fault-tolerance and low cos...
详细信息
ISBN:
(纸本)9780769533162
P2P networks have become a popular way to share large volumes of data due to its open and anonymous nature. P2P network model is designed with the targets of decentralization, well-scaling, fault-tolerance and low cost. Meanwhile however, new targets bring new challenges. P2P network expose more security threats due to its open and anonymous nature. P2P networks are vulnerable to attacks by malicious peers that have an incentive to spread viruses, in authentic data, or provide poor quality services. In this paper, we analyze the potential trust problems for different kinds of P2P networks, and then present the general trust factors that are indispensable for building trust models in any kinds of networks.
In order to build knowledge flow of textual topics, the enormous topics of scientific texts are needed to be managed and organized. Island is proposed to manage the topic of scientific text belonging to the same domai...
详细信息
In order to build knowledge flow of textual topics, the enormous topics of scientific texts are needed to be managed and organized. Island is proposed to manage the topic of scientific text belonging to the same domain. Interest Table is presented to store the textual topics that researcher is interested recently. KM-Chord is introduced to organize KM, which not only reflects the concepts (keywords) of textual topic, but also takes the relationships between concepts into account. So KM-Chord has more strong semantic information than traditional Chord. Knowledge Map (KM) representing textual topic can be located quickly with the help of KM-Chord, and the semantic information between KMs can be found (eg. similar and subtype relationship) for building the knowledge flow of the textual topics. The knowledge flow of textual topics can support knowledge innovation, cooperative teamwork, problem-solving and decision-making in e-Science Knowledge grid. Experiments show KM-Chord not only can locate and route the textual topic, but also build the knowledge flow of textual topic effectively. So the proposed KM-Chord is a good method to build the Knowledge Flow of textual topics for the e-science Knowledge grid, semanticgrid and current Web.
暂无评论