• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.2
• /
• pp.213-225
• /
• 2022

Containerization, a lightweight virtualization technology, enables agile deployments of enterprise-scale microservices in modern cloud environments. However, containerization also opens a new window for adversaries who aim to disrupt the cloud environments. Since microservices are composed of multiple containers connected through a virtual network, a single compromised container can carry out network-level attacks to hijack its neighboring containers. While existing solutions protect containers against such attacks by using network access controls, they still have severe limitations in terms of performance. More specifically, they significantly degrade network performance when processing packet payloads for L7 access controls (e.g., HTTP). To address this problem, we present BPFast, an eBPF/XDP-based payload inspection system for containers. BPFast inspects headers and payloads of packets at a kernel-level without any user-level components. We evaluate a prototype of BPFast on a Kubernetes environment. Our results show that BPFast outperforms state-of-the-art solutions by up to 7x in network latency and throughput.

• Journal of Information Management
• /
• v.43 no.3
• /
• pp.23-41
• /
• 2012

Citation analysis studies have reported many problems associated with data coverage problems common to popular citation databases such as Web of Science(WoS). In addition, the studies that analyzed citation patterns of Korean publications found that up to 75% of references in Korean publications were to international publications. As a first step in investigating the international coverage of WoS database, the study investigated the effect of adding citation data from Korea Science Citation Database(KSCD) to the impact factors and journal rankings of the journals listed in Journal Citation Reports. Specifically, the study mined the reference data from top 5 Korean Library and Information Science(KLIS) journals to recompute the impact factors reported in JCR 2009. Since the resulting journal rankings did not significantly differ from JCR 2009 rankings except for minor ranking changes, we analyzed additional citation data from 45 computer science and electrical engineering journals. Although the overall ranking difference was not statistically significant, one of the ranking partitions showed significant change. Such study findings despite its limited data sample suggest the potential impact of non-Western citation databases such as KSCD to bibliometric indicators provided by popular citation databases like WoS.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.7
• /
• pp.694-702
• /
• 2008

The web might be today's most widely-used technology and over the past few years, the web has evolved through various phases and is a mature technology Specially, the web interface technologies, such as folksonomy, tagging, RSS have been developed and changed with respect to user-friendly. In this paper, we should apply to not only the general structure scheme but also the tagging technology for POI information management and develop POIM(POI Management tool)for it. For POI service with user-friendly, the user can available to define, modify, and delete for a specific POI object based on XML by using POIM and then these should be saved on GML.

• The Journal of the Korea Contents Association
• /
• v.11 no.9
• /
• pp.515-523
• /
• 2011

It is important to manage large amount of information from various information providers for supporting recent information services such as providing global research trends, detecting emerging signal and listing leading researchers. For integrated management, definition of integrated metadata schema, data transformation and schema matching are needed. It is first necessary to analyze existing various metadata for defining integrated metadata schema. In this paper, we have analyzed several metadata of scientific journal papers by classifying semantics, content rules and syntax, and looked around considerations to make integrated schema or transform metadata. We have known that XML is used as a syntax for supporting convenience and various usage condition, and hierarchy element names and common elements in semantics are needed. We also have looked at elements having various content rules and related standards. We hope that this study will be used as basic research material of metadata integrated management, data transform and schema matching for interoperability.

• Journal of Information Management
• /
• v.39 no.4
• /
• pp.47-66
• /
• 2008

This study introduced the machine learning algorithms to overcome the many different limitations involved with manual classification and to provide the users with faster and more accurate classification service. The experiments objects of the study were consisted of 100 literature titles for each of the eight subject categories in MeSH. The algorithms used to the experiments included Neural network, C5.0, CHAID and KNN. As results, the combination of the neural network and C5.0 technique recorded classification accuracy of 83.75%, which was 2.5% and 3.75% higher than that of the neural network alone and C5.0 alone, respectively. The number represented the highest accuracy rates among the four classification experiments. Thus the use of the neural network and C5.0 technique together will result in higher accuracy rates than the techniques individually.

• Database Research
• /
• v.34 no.3
• /
• pp.86-98
• /
• 2018

Point-of-Interest(POI) recommendation systems suggest the most interesting POIs to users considering the current location and time. With the rapid development of smartphones, internet-of-things, and location-based social networks, it has become feasible to accumulate huge amounts of user POI visits. Therefore, instant recommendation of interesting POIs at a given time is being widely recognized as important. To increase the performance of POI recommendation systems, several studies extracting users' POI sequential preference from POI check-in data, which is intended for implicit feedback, have been suggested. However, when constructing a model utilizing sequential preference, the model encounters possibility of data distortion because of a low number of observed check-ins which is attributed to intensified data sparsity. This paper suggests refinement of temporal intervals based on data confidence. When building a POI recommendation system using temporal intervals to model the POI sequential preference of users, our methodology reduces potential data distortion in the dataset and thus increases the performance of the recommendation system. We verify our model's effectiveness through the evaluation with the Foursquare and Gowalla dataset.

• Journal of the Korean Society for information Management
• /
• v.28 no.2
• /
• pp.97-115
• /
• 2011

A large and reliable citation database is necessary to identify and analyze citation behavior of Korean researchers in science and technology. Korea Institute of Science and Technology Information (KISTI) built the Korea Science Citation Database (KSCD), and have provided Korea Science Citation Index (KSCI) and Korea Journal Citation Reports (KJCR) services. In this article, citing behavior of Korean scientists on Korean journals was examined by using the KSCD that covers 459 Korean core journals. This research dealt with (1) statistical numeric information of journals in KSCD, (2) analysis of document types cited, (3) ratio of domestic to international documents cited and ratio of citing different disciplines, (4) analysis on immediacy index, peak time, and half-life of cited documents, and (5) analysis on impact of journals based on KJCR citation indicators. From this research, we could find the immediacy citation rate (average 2.36%), peak-time (average 1.7 years) and half-life (average 5.2 years) of cited journals in Korea. We also found that the average journal self-citation rate is more than 50% in every field. In sum, citing behavior of Korean scientists on Korean journals was comprehensively identified from this research.

• Proceedings of the Korea Contents Association Conference
• /
• 2011.05a
• /
• pp.323-324
• /
• 2011

각 부처에서 관리되고 있는 국가R&D정보는 범부처 공동활용을 위하여 국가과학기술지식정보서비스(NTIS) 체계를 통하여 유통된다. 그러나 부처별 연구관리전문기관에서 NTIS로의 단방향성 R&D정보 유통체계로 인하여 부처별 전문기관에서 관리되는 R&D정보와 NTIS에서 관리되는 R&D정보간 데이터 불일치 문제가 발생한다. 그러므로 본 논문에서는 국가R&D정보의 관리현황 및 문제점을 분석하고, NTIS와 부처간 국가R&D정보의 동기화를 위한 국가R&D정보 유통체계 개선 프로세스를 제안한다.

• Korean Journal of Heritage: History & Science
• /
• v.57 no.1
• /
• pp.6-19
• /
• 2024

The Korean Archaeological Dictionary, which represents Korean archaeological knowledge information, contains refined and high-quality information written by expert collective intelligence. This is a characteristic that clearly distinguishes it from overseas archaeological data archives, and can be called differentiated infrastructure data. However, it has not played a role as an information service or knowledge information platform reflecting the latest digital technology. As a way to maximize these strengths and compensate for weaknesses, it was proposed to develop and operate a GIS-based knowledge and information platform for Korean archaeology. To realize this, it is necessary to develop a title management system centered on repositories and metadata that can collect and store various information, link open linked data design and related systems, develop a search function that can analyze and visualize data in response to the big data era, and establish a WebGIS-based information service system. This will be a platform to continuously manage, supplement, and update Korean archaeological knowledge information, build a ubiquitous environment where anyone can use information anytime, anywhere, and create various types of business models.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.1
• /
• pp.106-110
• /
• 2008

The requirement for customized learning environments is increasing with the development of Web 2.0 technology To personalize services in E-learning, there are various functions available to the user. However, some components have fixed or inflexible functionality that limits what can be changed, and customization information has not been used for other purposes. In this paper, we propose an e-Space manager for the dynamic customization of learning environment control units. The proposed method can confirm which areas users have customized and how they have customized them. The advantage is that users can control their own learning environments, including not only the format, but also the content within the limits of the underlying system. The customization information is captured to confirm the competency model.