Search

This article discusses the influence of semantics, knowledge, and advanced cyber-infrastructure on the development of computing technologies and systems. It highlights the challenges posed by the rapid development of Artificial Intelligence and Data Science. The research topics covered include semantics modeling, knowledge discovery in big data, advanced computing infrastructures, and applications in areas such as smart cities and healthcare. The article also presents four papers on link prediction, image captioning, combining graphics and uncertainty, and chatbots for psychotherapy. It concludes by discussing the emergence of a new research paradigm on semantics, knowledge, and cyberinfrastructure. [Extracted from the article]

The World Wide Web provides plentiful contents for Web-based learning, but its hyperlink-based architecture connects Web resources for browsing freely rather than for effective learning. To support effective learning, an e-learning system should be able to discover and make use of the semantic communities and the emerging semantic relations in a dynamic complex network of learning resources. Previous graph-based community discovery approaches are limited in ability to discover semantic communities. This paper first suggests the Semantic Link Network (SLN), a loosely coupled semantic data model that can semantically link resources and derive out implicit semantic links according to a set of relational reasoning rules. By studying the intrinsic relationship between semantic communities and the semantic space of SLN, approaches to discovering reasoning-constraint, rule-constraint, and classification-constraint semantic communities are proposed. Further, the approaches, principles, and strategies for discovering emerging semantics in dynamic SLNs are studied. The basic laws of the semantic link network motion are revealed for the first time. An e-learning environment incorporating the proposed approaches, principles, and strategies to support effective discovery and learning is suggested. Index Terms--Community discovery, e-learning, emerging semantics, semantic community, Semantic Link Network.

This paper presents a general virtual ring method to design and analyze small-world structured P2P networks on the base topologies embedded in ID spaces with distance metric. Its basic idea is to abstract a virtual ring from the base topology according to the distance metric, then build small-world long links in the virtual ring and map the links back onto the real network to construct the small-world routing tables for achieving logarithmic greedy routing. Four properties are proposed to characterize the base topologies that can be turned into small world by the virtual ring method. The virtual ring method is applied to the base topologies of d-torus with Manhattan distance, high-dimensional d-torus base topologies, and other base topologies including the unbalanced d- torus and the ring topology with tree distance. Theoretical analysis and experiments demonstrate the efficiency and resilience of the proposed routing schemas. Index Terms--P2P, small world, structured topology, routing table, query routing.

Establishing an appropriate semantic overlay on peer-to-peer (P2P) networks to obtain both semantic ability and scalability is a challenge. Current DHT-based P2P networks are limited in their ability to support a semantic search. This paper proposes the Distributed Suffix Tree (DST) overlay as the intermediate layer between the DHT overlay and the semantic overlay to support the search of a keyword sequence. Its time cost is sublinear with the length of the keyword sequence. Analysis and experiments show that the DST-based search is fast, load-balanced, and useful in realizing an accurate content search on P2P networks. Index Terms--DHT, knowledge grid, peer-to-peer, semantic overlay, suffix tree, load balance.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jss.2007.02.028 Byline: Hai Zhuge (a)(b)(d), Weiyu Guo (a)(b)(c) Keywords: Knowledge grid; Knowledge service; Knowledge flow network; Reputation; Market; Auction; Bargain Abstract: A knowledge service consists of systematic knowledge and the mechanism of using knowledge to perform a task. The supply of knowledge services forms a knowledge service layer over the knowledge flow network formed by free knowledge sharing. To stimulate the supply of knowledge services, this paper proposes a virtual knowledge service market by establishing reward and reputation mechanisms. Simulations demonstrate that a team with the market mechanism performs better than those without it. The virtual knowledge service market provides an experimental platform for exploring the rules of knowledge service such as the impact of individual behavior on states of individual and team as well as the change of the states. Author Affiliation: (a) China Knowledge Grid Research Group, Key Lab of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China (b) Graduate School of Chinese Academy of Sciences, Beijing, China (c) Shandong Institute of Light Industry, Jinan, Shandong, China (d) Hunan University, Changsha, Hunan, China Article History: Received 9 January 2007; Revised 11 February 2007; Accepted 12 February 2007 Article Note: (footnote) [star] This work was supported by National Basic Research Program of China (973 Project No. 2003CB317001), International Cooperation Project of MOST of China, the EU 6th Framework Project GREDIA (IST-FP6-034363), and the National Science Foundation of China (No. 60503047).

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jss.2006.03.001 Byline: Jie Liu (a)(b), Hai Zhuge (a) Keywords: Peer data management; Peer-to-peer; Resource space model; Semantic link network; Knowledge grid Abstract: This paper proposes a semantic-based P2P resource organization model R-Chord by incorporating the Resource Space Model (RSM), the P2P Semantic Link Network Model (P2PSLN) and the DHT Chord protocol. Peers provide services with each other according to the content of their resources and the related configuration information. It incorporates the classification semantics and the relational semantics to provide users and applications with a uniform view on distributed resources. Experiments show that, compared to the Chord approach, the R-Chord approach is more flexible to support semantic-based queries and can significantly decrease the average visiting number of and visiting times on peers for answering queries. Author Affiliation: (a) China Knowledge Grid Research Group, Key Lab of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, 100080, China (b) Graduate School of the Chinese Academy of Sciences, Beijing, 100080, China Article History: Received 15 December 2005; Revised 3 March 2006; Accepted 6 March 2006

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jss.2006.02.033 Byline: Hai Zhuge Keywords: Component; Model; Reasoning; Reuse; Rules; Knowledge grid; Service; Web Abstract: Model is a kind of codified knowledge that has been verified in solving problems. Solving a complex problem usually needs a set of models. Using components, the composition of a set of closely related models, could enhance the efficiency of solving problems as components are experience in form of knowledge network. Mappings between components vary with the relevancy of problems. Such mappings can help retrieve appropriate components when solving new problems. This paper investigates the relationship between components to identify these mappings by a set of semantic links and develops relevant rules to form a mechanism for semantically networking and using components. Flexible component reuse can be realized by semantic retrieval and rule reasoning. Extending the notion of component to a general service mechanism, this paper further investigates the organization, reuse and clustering of components in form of Web services and research spaces. The semantic component networking overlay supports intelligent e-Science Knowledge Grid applications by the synergy of static reuse and dynamic clustering of various research spaces and resources on-demand. Author Affiliation: Hunan Knowledge Grid Lab, Hunan University of Science and Technology, Hunan, China China Knowledge Grid Research Group, Key Lab of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences, P.O. Box 2704-28, Beijing 100080, China Article History: Received 5 January 2005; Revised 12 February 2006; Accepted 19 February 2006 Article Note: (footnote) [star] Research work was supported by the National Basic Research Program of China (973 project no. 2003CB317000) and the National Science Foundation of China.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jss.2005.08.022 Byline: Hai Zhuge, Xiangfeng Luo Keywords: Fuzzy cognitive map; Document understanding; Semantics; Knowledge Grid; Semantic Web Abstract: This paper proposes an approach to automatically generate semantics for scientific e-documents, and presents its applications in e-document understanding, question answering and question refinement. The approach uses not only keywords and their relations in e-documents, but also the implied meaning of co-occurred keywords that is hard to be exploited, represented and derived by previous semantic representation approaches. The proposed approach facilitates automatic construction, composition, decomposition and derivation of semantics at different granularity levels, which lay the basis for realizing intelligent services of the e-science Knowledge Grid. Author Affiliation: China Knowledge Grid Research Group, Key Lab of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences, P.O. Box 2704, Beijing 100080, PR China Article History: Received 24 November 2004; Revised 21 August 2005; Accepted 26 August 2005 Article Note: (footnote) [star] Research work is supported by the National Basic Research Program of China (973 project no. 2003CB317000) and the National Science Foundation of China (grants 60273020, 60402016 and 70271007).

The Knowledge Grid needs to operate with a scalable platform to provide large-scale intelligent services. A key function of such a platform is to efficiently support various complex queries in a dynamic large-scale network environment. This paper proposes a platform to support index-based path queries by incorporating a semantic overlay with an underlying structured P2P network that provides object location and management services. Various distributed indexing structures can be dynamically formed by publishing semantic objects as indexing nodes. Queries are forwarded along the chains of semantic object pointers to search for objects. We investigate the deployment of a scalable distributed trie index for broadcast queries on key strings, propose a decentralized load balancing method for solving the problem of uneven load distribution incurred by heterogeneity of loads and node capacities and by the distributed trie index, and give an approach for improving the availability of the semantic overlay and its trie index. Experiments demonstrate the scalability of the proposed platform. Index Terms--Peer-to-peer, semantic overlay, knowledge grid, path query, distributed trie index, load balancing, replication.

Summary: News image captioning aims to generate captions or descriptions for news images automatically, serving as draft captions for creating news image captions manually. News image captions are different from generic captions as news image captions contain more detailed information such as entity names and events. Therefore, both images on news and the accompanying text are the source of generating caption of news image. Pointer‐generator network is a neural method defined for text summarization. This article proposes the Multimodal pointer‐generation network by incorporating visual information into the original network for news image captioning. The multimodal attention mechanism is proposed by splitting attention into visual attention paid to the image and textual attention paid to the text. The multimodal pointer mechanism is proposed by using both textual attention and visual attention to compute pointer distributions, where visual attention is first transformed into textual attention via the word‐image relationships. The multimodal coverage mechanism is defined to reduce repetitions of attentions or repetitions of pointer distributions. Experiments on the DailyMail test dataset and the out‐of‐domain BBC test dataset show that the proposed model outperforms the original pointer‐generator network, the generic image captioning method, the extractive news image captioning method, and the LDA‐based method according BLEU, METEOR, and ROUGL‐L evaluations. Experiments also show that the proposed multimodal coverage mechanisms can improve the model, and that transforming visual attention to pointer distributions can improve the model. [ABSTRACT FROM AUTHOR]

The key to realize advanced document summarization is semantic representation of documents. This paper investigates the role of Semantic Link Network in representing and understanding documents for multi-document summarization. It proposes a novel abstractive multi-document summarization framework by first transforming documents into a Semantic Link Network of concepts and events and then transforming the Semantic Link Network into the summary of the documents based on the selection of important concepts and events while keeping semantics coherence. Experiments on benchmark datasets show that the proposed summarization approach significantly outperforms relevant state-of-the-art baselines and the Semantic Link Network plays an important role in representing and understanding documents. [ABSTRACT FROM AUTHOR]