Search results

Malicious attacks extremely traumatize knowledge networks due to increasing interdependence among knowledge elements. Therefore, exposing the damage of malicious attacks to knowledge networks has important theoretical and practical significance. Despite the insights being offered by the growing research stream, few studies discuss the diverse responses of knowledge networks’ robustness to different target-attacks, and the authors lack sufficient knowledge of which forms of malicious attacks constitute greater disaster when knowledge networks evolve to different stages. Given the irreversible consequences of malicious attacks on knowledge networks, this paper aims to examine the impacts of different malicious attacks on the robustness of knowledge networks.

On the basic of dividing malicious attacks into six forms, the authors incorporate two important aspects of robustness of knowledge networks – structure and function – in a research framework, and use maximal connected sub-graphs and network efficiency, respectively, to measure structural and functional robustness. Furthermore, the authors conceptualize knowledge as a multi-dimensional structure to reflect the heterogeneous nature of knowledge elements, and design the fundamental rules of simulation. NetLogo is used to simulate the features of knowledge networks and their changes of robustness as they face different malicious attacks.

First, knowledge networks gradually form more associative integrated structures with evolutionary progress. Second, various properties of knowledge elements play diverse roles in mitigating damage from malicious attacks. Recalculated-degree-based attacks cause greater damage than degree-based attacks, and structure of knowledge networks has higher resilience against ability than function. Third, structural robustness is mainly affected by the potential combinatorial value of high-degree knowledge elements, and the combinatorial potential of high-out-degree knowledge elements. Forth, the number of high in-degree knowledge elements with heterogeneous contents, and the inverted U-sharp effect contributed by high out-degree knowledge elements are the main influencers of functional robustness.

The authors use the frontier method to expose the detriments of malicious attacks both to structural and functional robustness in each evolutionary stage, and the authors reveal the relationship and effects of knowledge-based connections and knowledge combinatorial opportunities that contribute to maintaining them. Furthermore, the authors identify latent critical factors that may improve the structural and functional robustness of knowledge networks.

First, from the dynamic evolutionary perspective, the authors systematically examine structural and functional robustness to reveal the roles of the properties of knowledge element, and knowledge associations to maintain the robustness of knowledge networks. Second, the authors compare the damage of six forms of malicious attacks to identify the reasons for increased robustness vulnerability. Third, the authors construct the stock, power, expertise knowledge structure to overcome the difficulty of knowledge conceptualization. The results respond to multiple calls from different studies and extend the literature in multiple research domains.

To simulate large parasitic resistive networks, one must reduce the size of the circuit models through methods that are accurate and preserve terminal connectivity and network sparsity. The purpose here is to present such a method, which exploits concepts from graph theory in a systematic fashion.

The model order reduction problem is formulated for parasitic resistive networks through graph theory concepts and algorithms are presented based on the notion of vertex cut in order to reduce the size of electronic circuit models. Four variants of the basic method are proposed and their respective merits discussed.

The algorithms proposed enable the production of networks that are significantly smaller than those produced by earlier methods, in particular the method described in the report by Lenaers entitled “Model order reduction for large resistive networks”. The reduction in the number of resistors achieved through the algorithms is even more pronounced in the case of large networks.

The paper seems to be the first to make a systematic use of vertex cuts in order to reduce a parasitic resistive network.

The purpose of this study is to develop a novel region-based convolutional neural networks (R-CNN) approach that is more efficient while at least as accurate as existing R-CNN methods. In this way, the proposed method, namely R²-CNN, provides a more powerful tool for pedestrian extraction for person re-identification, which involve a huge number of images and pedestrian needs to be extracted efficiently to meet the real-time requirement.

The proposed R²-CNN is tested on two types of data sets. The first one the USC Pedestrian Detection data set, which consists of three sub-sets USC-A, UCS-B and USC-C, with respect to their characteristics. This data set is used to test the performance of R²-CNN in the pedestrian extraction task. The speed and performance of the investigated algorithms were collected. The second data set is the PASCAL VOC 2007 data set, which is a common benchmark data set for object detection. This data set was used to analyze characteristics of R²-CNN in the case of general object detection task.

This study proposes a novel R-CNN method that is both more efficient and more accurate than existing methods. The method, when used as an object detector, would facilitate the data preprocessing stage of person re-identification.

The study proposes a novel approach for object detection, which shows advantages in both efficiency and accuracy for pedestrian detection task. It contributes to both data preprocessing for person re-identification and the research on deep learning.

The purpose of this paper is to examine the evolution of collaboration among researchers in Library Hi Tech based on the co-authorship network analysis.

The Library Hi Tech publications were retrieved from Web of Science database between 2006 and 2017. Social network analysis based on co-authorship was analyzed by using BibExcel software and a visual knowledge map was generated by Pajek. Three important social capital indicators: degree centrality, closeness centrality and betweenness centrality were calculated to indicate the co-authorship. Cohesive subgroup analysis which includes components and k-core was then applied to show the connectivity of co-authorship network of Library Hi Tech.

The results indicated that around 42 percent of the articles were written by single author, while an increasing trend of multi-authored articles suggesting the collaboration among researchers in librarian research field becomes popular. Furthermore, the social network analysis identified authorship network with three core authors – Markey, K., Fourie, I. and Li, X. Finally, six core subgroups each included six or seven tightly connected researchers were also identified.

This study contributed to the existing literature by revealing the co-authorship network in librarian research field. Key researchers in the major subgroup were identified. This is one of the limited studies that describe the collaboration network among authors from different perspectives showing a more comprehensive co-authorship network.

The application of network theory and social network analysis (SNA) to tourism and hospitality is recent. Nonetheless, several authors have been applying the method contributing to regional planning, local-level tourism networks, tourism policy and governance, innovation, entrepreneurship, knowledge transfer, and learning. This chapter aims to characterize the use of SNA in tourism and hospitality research. Specifically, it intends to: (i) present the framework of SNA in a methodological perspective; (ii) perform a bibliometric analysis of SNA use in tourism and hospitality research; (iii) systematize the dimensions and metrics that researchers can use to apply SNA, namely the relevance for tourism; and (iv) present a case study analyzing tourism innovation networks. This chapter brings important contributions to tourism and hospitality research and practice, by focusing on the theoretical framework and practical application of SNA, providing relevant conceptual and practical knowledge that will empower researchers to use this method in tourism and hospitality studies.

As most existing computer‐aided design systems do not provide part feature information which is essential for process planning, automatic part feature recognition systems serve as an important link between Computer Aided Design (CAD) and Computer Aided Process Planning (CAPP). Attempts to provide a structural framework for understanding various issues related to part feature recognition. Reviews previous work in the field of part feature recognition and classifies known feature recognition systems for the sake of updating information and future research. Briefly introduces about 12 systems. Studies 31 systems and lists them in the Appendix based on 60 references. Comments on future research directions.

Purchase conversion prediction aims to improve user experience and convert visitors into real buyers to drive sales of firms; however, the total conversion rate is low, especially for e-retailers. To date, little is known about how e-retailers can scientifically detect users' intents within a purchase conversion funnel during their ongoing sessions and strategically optimize real-time marketing tactics corresponding to dynamic intent states. This study mainly aims to detect a real-time state of the conversion funnel based on graph theory, which refers to a five-class classification problem in the overt real-time choice decisions (RTCDs)—click, tag-to-wishlist, add-to-cart, remove-from-cart and purchase—during an ongoing session.

The authors propose a novel graph-theoretic framework to detect different states of the conversion funnel by identifying a user's unobserved mindset revealed from their navigation process graph, namely clickstream graph. First, the raw clickstream data are identified into individual sessions based on a 30-min time-out heuristic approach. Then, the authors convert each session into a sequence of temporal item-level clickstream graphs and conduct a temporal graph feature engineering according to the basic, single-, dyadic- and triadic-node and global characteristics. Furthermore, the synthetic minority oversampling technique is adopted to address with the problem of classifying imbalanced data. Finally, the authors train and test the proposed approach with several popular artificial intelligence algorithms.

The graph-theoretic approach validates that users' latent intent states within the conversion funnel can be interpreted as time-varying natures of their online graph footprints. In particular, the experimental results indicate that the graph-theoretic feature-oriented models achieve a substantial improvement of over 27% in line with the macro-average and micro-average area under the precision-recall curve, as compared to the conventional ones. In addition, the top five informative graph features for RTCDs are found to be Transitivity, Edge, Node, Degree and Reciprocity. In view of interpretability, the basic, single-, dyadic- and triadic-node and global characteristics of clickstream graphs have their specific advantages.

The findings suggest that the temporal graph-theoretic approach can form an efficient and powerful AI-based real-time intent detecting decision-support system. Different levels of graph features have their specific interpretability on RTCDs from the perspectives of consumer behavior and psychology, which provides a theoretical basis for the design of computer information systems and the optimization of the ongoing session intervention or recommendation in e-commerce.

To the best of the authors' knowledge, this is the first study to apply clickstream graphs and real-time decision choices in conversion prediction and detection. Most studies have only meditated on a binary classification problem, while this study applies a graph-theoretic approach in a five-class classification problem. In addition, this study constructs temporal item-level graphs to represent the original structure of clickstream session data based on graph theory. The time-varying characteristics of the proposed approach enhance the performance of purchase conversion detection during an ongoing session.

This study aims to extract knowledge of ancient Chinese scientific and technological documents bibliographic summaries (STDBS) and provide the knowledge graph (KG) comprehensively and systematically. By presenting the relationship among content, discipline, and author, this study focuses on providing services for knowledge discovery of ancient Chinese scientific and technological documents.

This study compiles ancient Chinese STDBS and designs a knowledge mining and graph visualization framework. The authors define the summaries' entities, attributes, and relationships for knowledge representation, use deep learning techniques such as BERT-BiLSTM-CRF models and rules for knowledge extraction, unify the representation of entities for knowledge fusion, and use Neo4j and other visualization techniques for KG construction and application. This study presents the generation, distribution, and evolution of ancient Chinese agricultural scientific and technological knowledge in visualization graphs.

The knowledge mining and graph visualization framework is feasible and effective. The BERT-BiLSTM-CRF model has domain adaptability and accuracy. The knowledge generation of ancient Chinese agricultural scientific and technological documents has distinctive time features. The knowledge distribution is uneven and concentrated, mainly concentrated on C1-Planting and cultivation, C2-Silkworm, and C3-Mulberry and water conservancy. The knowledge evolution is apparent, and differentiation and integration coexist.

This study is the first to visually present the knowledge connotation and association of ancient Chinese STDBS. It solves the problems of the lack of in-depth knowledge mining and connotation visualization of ancient Chinese STDBS.

Siphon-based deadlock control in a flexible manufacturing system (FMS) suffers from the problems of computational and structural complexity since the number of siphons grows exponentially with respect to the size of its Petri net model. In order to reduce structural complexity of a supervisor, a set of elementary siphons derived from all strict minimal siphons (SMS) is explicitly controlled. The purpose of this paper is through fully investigating the structure of a class of generalized Petri nets, WS3PR, to compute all SMS and a compact set of elementary siphons.

Based on graph theory, the concepts of initial resource weighted digraphs and restricted subgraphs are proposed. Moreover, the concept of augmented siphons is proposed to extend the application of elementary siphons theory for WS3PR. Consequently, the set of elementary siphons obtained by the proposed method is more compact and well suits for WS3PR.

In order to demonstrate the proposed method, an FMS example is presented. All SMS and elementary siphons can be derived from initial resource weighted digraphs. Compared with those obtained by the method in Li and Zhou, the presented method is more effective to design a structural simple liveness-enforcing supervisor for WS3PR.

This work presents an effective method of computing SMS and elementary siphons for WS3PR. Monitors are added for the elementary siphons only, and the controllability of every dependent siphon is ensured by properly supervising its elementary ones. A same set of elementary siphons can be admitted by different WS3PR with isomorphic structures.

Design is a time-consuming process for mechanical production. Some design structures frequently occur in different products and can be shared by multiple assembly models. Thus, identifying these structures and adding them to a design knowledge library significantly speed up the design process. Most studies addressing this issue have traditionally focused on part models and have not extended to assembly models. This paper aims to find a method for common design structure discovery in assembly models.

Computer-aided design models have a great deal of valuable information defined by different designers in the design stages, especially the assembly models, which are actually carriers of information from multiple sources. In this paper, an approach for discovering a common design structure in assembly models is proposed by comparing information from multiple sources. Assembly models are first represented as attribute connection graphs (ACGs), in which we mainly consider topological information and various attributes of parts and connections. Then, we apply a K-means clustering method based on a similarity analysis of different attributes to classify the parts and connections and transform ACGs of assemblies into type code graphs (TCGs). After this, a discovery algorithm that improves upon fast frequent subgraph mining is used to identify common design structures in assemblies.

A new method was developed for discovering common design structures in assembly models, considering the similarity of information from multiple sources and allowing some differences in the details to keep both commonalities and individualities of common design structures.

Experiments show that the proposed method is efficient and can produce a reasonable result.

This discovery method helps designers find common design structures from different assembly models and shorten the design cycle. It is an effective approach to build a knowledge library for product design that can shorten the design cycle.