• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.12
• /
• pp.485-496
• /
• 2018

So far, the deep learning, a field of artificial intelligence, has achieved remarkable results in solving problems from unstructured data. However, it is difficult to comprehensively judge situations like humans, and did not reach the level of intelligence that deduced their relations and predicted the next situation. Recently, deep neural networks show that artificial intelligence can possess powerful relational reasoning that is core intellectual ability of human being. In this paper, to analyze and observe the performance of Relation Networks (RN) among the neural networks for relational reasoning, two types of RN-based deep neural network models were constructed and compared with the baseline model. One is a visual question answering RN model using Sort-of-CLEVR and the other is a text-based question answering RN model using bAbI task. In order to maximize the performance of the RN-based model, various performance improvement experiments such as hyper parameters tuning have been proposed and performed. The effectiveness of the proposed performance improvement methods has been verified by applying to the visual QA RN model and the text-based QA RN model, and the new domain model using the dialogue-based LL dataset. As a result of the various experiments, it is found that the initial learning rate is a key factor in determining the performance of the model in both types of RN models. We have observed that the optimal initial learning rate setting found by the proposed random search method can improve the performance of the model up to 99.8%.

• Korean Journal of Social Welfare
• /
• v.61 no.3
• /
• pp.283-305
• /
• 2009

The purpose of this study is to develop a scale for the social network of college students in Korea. The social network scale for college students in this paper was developed through 1) literature review and item development 2) focus-group meeting 3) depth-interview with college students 4) pilot-studies and 5) study. The scale of 23 items has been constructed with three features : The first is 'relational feature' with 2 items; the second 'structural feature' with 7 items; and the third 'functional feature' with 14 items. The functional feature in this social network scale has consisted of emotional, economic, spiritual, and socio-relational aspects and has been revealed to the validity and the reliability of Cronbach's ${\alpha}$ 0.716. This study would contribute for college students to solve the many problems in their own lives through the social networks.

• Journal of the Ergonomics Society of Korea
• /
• v.34 no.5
• /
• pp.377-399
• /
• 2015

Objective: The aim of this study is to understand and identify the critical issues in vision research area using content analysis and network analysis. Background: Vision, the most influential factor in information processing, has been studied in a wide range of area. As studies on vision are dispersed across a broad area of research and the number of published researches is ever increasing, a bibliometric analysis towards literature would assist researchers in understanding and identifying critical issues in their research. Method: In this study, content and network analysis were applied on the meta-data of literatures collected using three search keywords: 'visual search', 'eye movement', and 'eye tracking'. Results: Content analysis focuses on extracting meaningful information from the text, deducting seven categories of research area; 'stimuli and task', 'condition', 'measures', 'participants', 'eye movement behavior', 'biological system', and 'cognitive process'. Network analysis extracts relational aspect of research areas, presenting characteristics of sub-groups identified by community detection algorithm. Conclusion: Using these methods, studies on vision were quantitatively analyzed and the results helped understand the overall relation between concepts and keywords. Application: The results of this study suggests that the use of content and network analysis helps identifying not only trends of specific research areas but also the relational aspects of each research issue while minimizing researchers' bias. Moreover, the investigated structural relationship would help identify the interrelated subjects from a macroscopic view.

• Journal of Digital Convergence
• /
• v.11 no.11
• /
• pp.209-217
• /
• 2013

This study focuses on the analysis of purchased items' relationship occurred by consumers' purchasing behavior observed in on-line shopping mall based on social network analysis. In order to find relational characteristics of each item for establishing marketing strategy, we apply three definitions of centrality in network, which are degree centrality, betweenness centrality, and closeness centrality in the purchased items' network. Thus, the research results provide the criteria for selection of market segmentation variables. Furthermore, the details of case has been introduced to validate the analyzed results in terms of marketing strategy, and supporting evidences are provided accordingly.

• Korean Management Science Review
• /
• v.32 no.4
• /
• pp.81-96
• /
• 2015

This paper models knowledge diffusion on an inter-organizational network. Based on literatures related to knowledge diffusion, the model considers critical factors that affect diffusion behavior including nodal property, relational property, and environmental property. We examine the relationships among network structure, knowledge search, and diffusion performance. Through a massive simulation runs based on the agent-based model, we find that the average path length of a network decreases a firm's cumulative knowledge stock, whereas the clustering coefficient of a firm has no significant relationship with the firm's knowledge. We also find that there is an interaction effect of network structure and the range of knowledge search on knowledge diffusion. Specifically, in a network of a larger average path length (APL) the marginal effect of search conduct is significantly greater than in that of a smaller APL.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.6 no.4
• /
• pp.37-50
• /
• 2003

In the paper, we describe the query extension techniques for spatiotemporal query functionalities on object-relational DBMS. The spatial objects in real world change the shapes over time. Spatiotemporal databases support to manage a temporal dimension as well as a spatial dimension for history of the objects. The proposed techniques can make conventional object-relational databases to support spatiotemporal databases system by the implementation and inheritance of abstract data types. We define and implement spatial and temporal classes as superclass. And, spatiotemporal classes inherits and extends the classes. The proposed extensions make it easy that conventional database systems not only are transformed into the spatiotemporal database systems, but also do not need to be changed to support spatiotemporal applications.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.3 no.3
• /
• pp.90-97
• /
• 2000

The study was aimed to extract the improvement measures for the problem on GIS application GIS for water supply and sewerage pipeline facility management in the Jechon city. For this, it performed of analysis of their working and modeling with other relational contents of the water and sewer facility management. As the results, the implementation of water and sewer facility management system by use of GIS has to applying development through relational analysis not only pipeline facility and leaking water protection, pipeline network analysis but also digital topography, drawing data, water user's information.

• 한국HCI학회:학술대회논문집
• /
• 2006.02b
• /
• pp.733-743
• /
• 2006

블로그는 현재 광범위하게 대중성을 얻고 있는 온라인 커뮤니케이션의 최근 형태이며, 블로그 사용이 빠른 속도로 주류를 형성해 나가고 있다. 초기 HCI 분야에서의 온라인 커뮤니케이션(CMC)에 대한 연구는 사용자가 시스템의 소프트웨어나 하드웨어의 인터페이스를 어떻게 대하는가 하는 점에 초점이 맞춰졌었지만, 최근에는 온라인 커뮤니티, 블로그처럼 사용자들이 인터넷을 통한 사회화가 어느 때보다 활발하게 이루어지고 있기 때문에, 시스템의 사용성에 대한 고려만으로는 충분하지가 않다. 즉, 시스템이 어떻게 사용자들 간에 사회적 상호 교류를 지원할 수 있고, 이러한 사회적 교류가 원활하게 이루어지기 위해서 시스템이 어떻게 설계되어야 하는 지에 대한 이해가 필요하다. 따라서 본 연구에서는 실제 블로그 이용자들을 대상으로 이들과 상호 연결된 블로거들과의 온라인에서의 관계적 행위를 살펴보고, 또한 이러한 행위가 네트워크 구조적 특성과는 어떠한 상관 관계를 가지는지를 심층 인터뷰와 사회 네트워크 분석법을 이용하여 탐색적으로 알아보고자 한다. 본 연구 결과를 통해서 얻은 사용자들의 온라인 교류 시 보이는 행위적 특성을 바탕으로 온라인 개인 미디어 서비스 설계에 대한 가이드라인을 제안해 보고자 한다.

• Korean journal of communication and information
• /
• v.60
• /
• pp.75-103
• /
• 2012

The purpose of this study is to explore the meaning and the effect of communication as social capital, which needs to be evaluated empirically focusing on corporate reputation. Also, it tried to analyze consumers' communication network in the structural, substantial, and relational level, which is to verify how characteristics and meanings of communication network structure affect to a good corporate reputation. A survey toward 200 participants was conducted during 5 days from March 29 to April 3, 2012. Characteristics of communication network structure of a corporation with higher reputation is analyzed using the index such as degree, degree centrality, and density. The findings of the study show that a corporation with higher reputation has higher network degree, degree centrality, and density compared to a corporation with lower reputation. Consumers of a corporation with higher reputation get information from various overlapping sources. It allows them to share similar interpretation, which could elevate the degree, degree centrality, and density of network. It also proved that when the network density is high, a corporation with higher reputation can distribute information much faster and easier. Moreover, in the substantial level of social capital, product/service information network has high degree and density rather than corporate issue information network. Likewise, degree and density of information acquisition network was higher than those of information provision network. Also, this study verified the effect and relationship between the network structure characteristics and corporate loyalty in a relational level. In this way, the positive effect of the degree centrality on corporate loyalty was supported. In conclusion, as consumers share more information from overlapping sources, the degree of communication network gets higher. Throughout this network, the diffusion of information among consumers would be activated, and this confirmed that corporate reputation and corporate loyalty is closely related.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.9
• /
• pp.76-84
• /
• 2012

Visualizations have played an important role in understanding new insights of users of social network for social network analysis. Most of the previous works of visualization focus on representing user's relationship on social network by a complex multi dimension graph. However, this method is difficult to identify the important of relationship to focus on personal user intuitively. Besides, the user's messages to reflect the interrelation between users is insufficient, since most of visualization methods represent the user relationship using information of interaction between nodes on networks. In order to resolve above problem, this paper proposes a new visualization method to visualize user based hierarchy that uses internal relationship of users by fuzzy relational product and external access information of network.