• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.2030-2052
• /
• 2023

Analysis of a bipartite (two-mode) network is a significant research area to understand the formation of social communities, economic systems, drug side effect topology, etc. in complex information systems. Most of the previous works talk about a projection-based model or latent feature model, which predicts the link based on singular similarity. The projection-based models suffer from the loss of structural information in the projected network and the latent feature is hardly present. This work proposes a novel method for link prediction in the bipartite network based on an ensemble of composite similarities, overcoming the issues of model-based and latent feature models. The proposed method analyzes the structure, neighborhood nodes as well as latent attributes between the nodes to predict the link in the network. To illustrate the proposed method, experiments are performed with five real-world data sets and compared with various state-of-art link prediction methods and it is inferred that this method outperforms with ~3% to ~9% higher using area under the precision-recall curve (AUC-PR) measure. This work holds great significance in the study of biological networks, e-commerce networks, complex web-based systems, networks of drug binding, enzyme protein, and other related networks in understanding the formation of such complex networks. Further, this study helps in link prediction and its usability for different purposes ranging from building intelligent systems to providing services in big data and web-based systems.

• Genomics & Informatics
• /
• v.1 no.1
• /
• pp.1-6
• /
• 2003

In conclusion, the seemingly fuzzy and disorganized data of biology with thousands of different layers ranging from molecule to the Internet have refused so far to be mapped precisely and predicted successfully by mathematicians, physicists or computer scientists. Genomics and bioinformatics are the fields that process such complex data. The insights on the nature of biological entities as complex interaction networks are opening a door toward a generalization of the representation of biological entities. The main challenge of genomics and bioinformatics now lies in 1) how to data mine the networks of the domains of bioinformatics, namely, the literature, metabolic pathways, and proteome and structures, in terms of interaction; and 2) how to generalize the networks in order to integrate the information into computable genomic data for computers regardless of the levels of layer. Once bioinformatists succeed to find a general principle on the way components interact each other to form any organic interaction network at genomic scale, true simulation and prediction of life in silico will be possible.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.43 no.4
• /
• pp.543-552
• /
• 1994

This paper presents a decentralized quadratic regulation architecture with feedforward neural networks for the control problem of complex systems. In this method, the decentralized technique was used to treat several simple subsystems instead of a full complex system in order to reduce training time of neural networks, and the neural networks' nonlinear mapping ability is exploited to handle the nonlinear interaction variables between subsystems. The decentralized regulating architecture is composed of local neuro-controllers, local neuro-identifiers and an overall interaction neuro-identifier. With the interaction neuro-identifier that catches interaction characteristics, a local neuro-identifier is trained to simulate a subsystem dynamics. A local neuro-controller is trained to learn how to control the subsystem by using generalized Backprogation Through Time(BTT) algorithm. The proposed neural network based decentralized regulating scheme is applied in the power System Stabilization(PSS) control problem for an imterconnected power system, and compared with that by a conventional centralized LQ regulator for the power system.

• Structural Monitoring and Maintenance
• /
• v.5 no.4
• /
• pp.445-461
• /
• 2018

The aim of this research is to evaluate the effects of the number of spans, height of spans, number of floors, height of floors, column to beam moment of inertia ratio, and plastic joints distance of beams from columns on the ductility of moment frames. For the facility in controlling the ductility of the frames, this paper offers a simple relation instead of complex equations of different codes. For this purpose, 500 analyzed and designed frames were randomly selected, and their ductility was calculated by the use of nonlinear static analysis. The results cleared that the column-to-beam moment of inertia ratio had the highest effect on ductility, and if this relation was more than 2.8, there would be no need for using the complex relations of codes for controlling the ductility of frames. Finally, the ductility of the most frames of this research could be estimated by using the combination of genetic algorithm and artificial neural networks properly.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.6
• /
• pp.2979-2995
• /
• 2017

In recent years, there has been an increasing number of studies focused on identifying a set of spreaders to maximize the influence of spreading in complex networks. Although the k-core decomposition can effectively identify the single most influential spreader, selecting a group of nodes that has the largest k-core value as the seeds cannot increase the performance of the influence maximization because the propagation sphere of this group of nodes is overlapped. To overcome this limitation, we propose a neighborhood coreness cover and discount heuristic algorithm named "NCCDH" to identify a set of influential and decentralized seeds. Using this method, a node in the high-order shell with the largest neighborhood coreness and an uncovered status will be selected as the seed in each turn. In addition, the neighbors within the same shell layer of this seed will be covered, and the neighborhood coreness of the neighbors outside the shell layer will be discounted in the subsequent round. The experimental results show that with increases in the spreading probability, the NCCDH outperforms other algorithms in terms of the affected scale and spreading speed under the Susceptible-Infected-Recovered (SIR) and Susceptible-Infected (SI) models. Furthermore, this approach has a superior running time.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.12 no.3
• /
• pp.1-9
• /
• 2016

The current meaning of "Big Data" refers to all the techniques for value eduction and actionable analytics as well management tools. Particularly, with the advances of wireless sensor networks, they yield diverse patterns of digital records. The records are mostly semi-structured and unstructured data which are usually beyond of capabilities of the management tools. Such data are rapidly growing due to their complex data structures. The complex type effectively supports data exchangeability and heterogeneity and that is the main reason their volumes are getting bigger in the sensor networks. However, there are many errors and problems in applications because the managing solutions for the complex data model are rarely presented in current big data environments. To solve such problems and show our differentiation, we aim to provide the solution of actionable analytics and semantic reusability in the sensor web based streaming big data with new data structure, and to empower the competitiveness.

• International Journal of Computer Science & Network Security
• /
• v.21 no.12spc
• /
• pp.643-647
• /
• 2021

The use of data analysis on large volumes of data constitutes a challenge for deducting knowledge and new information. Data can be heterogeneous and complex: Semi-structured data (Example: XML), Data from social networks (Example: Tweets) and Factual data (Example: Spreading of Covid-19). In this paper, we propose a generic multidimensional model in order to analyze complex data, according to several dimensions.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.16 no.3
• /
• pp.157-168
• /
• 2014

From a complex systems perspective, ecohydrological systems in forests may be characterized with (1) large networks of components which give rise to complex collective behaviors, (2) sophisticated information processing, and (3) adaptation through self-organization and learning processes. In order to demonstrate such characteristics, we applied the recently proposed 'process networks' approach to a temperate deciduous forest in Gwangneung National Arboretum in Korea. The process network analysis clearly delineated the forest ecohydrological systems as the hierarchical networks of information flows and feedback loops with various time scales among different variables. Several subsystems were identified such as synoptic subsystem (SS), atmospheric boundary layer subsystem (ABLS), biophysical subsystem (BPS), and biophysicochemical subsystem (BPCS). These subsystems were assembled/disassembled through the couplings/decouplings of feedback loops to form/deform newly aggregated subsystems (e.g., regional subsystem) - an evidence for self-organizing processes of a complex system. Our results imply that, despite natural and human disturbances, ecosystems grow and develop through self-organization while maintaining dynamic equilibrium, thereby continuously adapting to environmental changes. Ecosystem integrity is preserved when the system's self-organizing processes are preserved, something that happens naturally if we maintain the context for self-organization. From this perspective, the process networks approach makes sense.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.8
• /
• pp.2587-2605
• /
• 2022

In order to efficiently detect community structure in complex networks, community detection algorithms can be designed from the perspective of node similarity. However, the appropriate parameters should be chosen to achieve community division, furthermore, these existing algorithms based on the similarity of common neighbors have low discrimination between node pairs. To solve the above problems, a noval community detection algorithm using closeness similarity based on common neighbor node clustering entropy is proposed, shorted as CSCDA. Firstly, to improve detection accuracy, common neighbors and clustering coefficient are combined in the form of entropy, then a new closeness similarity measure is proposed. Through the designed similarity measure, the closeness similar node set of each node can be further accurately identified. Secondly, to reduce the randomness of the community detection result, based on the closeness similar node set, the node leadership is used to determine the most closeness similar first-order neighbor node for merging to create the initial communities. Thirdly, for the difficult problem of parameter selection in existing algorithms, the merging of two levels is used to iteratively detect the final communities with the idea of modularity optimization. Finally, experiments show that the normalized mutual information values are increased by an average of 8.06% and 5.94% on two scales of synthetic networks and real-world networks with real communities, and modularity is increased by an average of 0.80% on the real-world networks without real communities.

• Journal of the Korean Geographical Society
• /
• v.37 no.2
• /
• pp.111-130
• /
• 2002

Major purposes of this stud? are to analyze Korean firms'innovation networks and sources of knowledge for innovation and to understand their spatial dimensions. In the innovation networks, parent firms are most important for subcontracting firms, while suppliers, customers and competitors are relatively important for independent firms. However, in the future innovation networks, it is expected that government-sponsored research institutions and university wilt become more important on the one hand, networks with foreign firms will become more important on the other hand. Regarding the process of innovation, distance does not matter for the acquisition of codified knowledge. Spatial proximity is, however, critical for the acquisition of tacit knowledge because discussions and researches in a research division within a firm, personal networks of CEO and workers who are responsible for innovation activity, and inter-firm relations with suppliers and customer in a region are regarded important as sources of tacit knowledge. Overall, the innovation networks are different between the Capital Region and non-Capital Region as well as between the industrial complex and non-industrial complex, suggesting that different regional innovation strategies and policies should be established and implemented by considering such regional specificities. Finally, based on the results of this study several policy implications are suggested.