• Journal of Korean Society for Geospatial Information Science
• /
• v.13 no.4 s.34
• /
• pp.27-32
• /
• 2005

DEM is needed for urban modeling, forecasting of floods and the analysis of slope and aspect. It has been generated using digital maps, aerial photos or satellite imageries. Recently, however, many studies on DEM generation from LiDAR data has been conducted because of its efficiency and accuracy. Filtering is said to be the process of making DEM by eliminating non-ground points from LiDAR data. In most researches, some input parameters such as the size of filter are required. The purpose of this investigation is to automatically obtain DEM by eliminating objects of various sizes without the knowledge of the objects' sizes. The experimental results show that most of objects on steep terrain are eliminated by the proposed method.

• Journal of Intelligence and Information Systems
• /
• v.8 no.2
• /
• pp.139-157
• /
• 2002

The rapid growth of e-commerce has made both companies and customers face a new situation. Whereas companies have become to be harder to survive due to more and more competitions, the opportunity for customers to choose among more and more products has increased. So, the recommender systems that recommend suitable products to the customer have an important position in E-commerce. This research introduces collaborative filtering based recommender system which helps customers find the products they would like to purchase by producing a list of top-N recommended products. The suggested methodology is based on decision tree, product taxonomy, and association rule mining. Decision tree is used to select target customers, who have high possibility of purchasing recommended products. We applied the recommender system to a Korean department store. The methodology is evaluated with the analysis of a real department store case and is compared with other methodologies.

• The Journal of the Korea Contents Association
• /
• v.8 no.2
• /
• pp.1-8
• /
• 2008

The information system in which users can utilize to control and to get the filtered information efficiently has appeared. Content-based filtering can reflect content information, and it provides recommendation by comparing the feature information about item and the profile of preference. This has the shortcoming of the varying accuracy of prediction depending on teaming method. This paper suggests the discovery of preference through learning the profile for the content-based filtering. This study improves the accuracy of recommendation through learning the profile according to granting the preference of 6 levels to estimated value in order to solve the problem. Finally, to evaluate the performance of the proposed method, this study applies to MovieLens dataset, and it is compared with the performance of previous studies.

• Journal of Broadcast Engineering
• /
• v.17 no.5
• /
• pp.742-755
• /
• 2012

In this paper, we propose a parallel deblocking algorithm to resolve workload imbalance when the deblocking filter of high efficiency video coding (HEVC) decoder is parallelized. In HEVC, the deblocking filter which is one of the in-loop filters conducts two-step filtering on vertical edges first and horizontal edges later. The deblocking filtering can be conducted with high-speed through data-level parallelism because there is no dependency between adjacent edges for deblocking filtering processes. However, workloads would be imbalanced among regions even though the same amount of data for each region is allocated, which causes performance loss of decoder parallelization. In this paper, we solve the problem for workload imbalance by predicting the complexity of deblocking filtering with coding unit (CU) depth information at a coding tree block (CTB) and by allocating the same amount of workload to each core. Experimental results show that the proposed method achieves average time saving (ATS) by 64.3%, compared to single core-based deblocking filtering and also achieves ATS by 6.7% on average and 13.5% on maximum, compared to the conventional uniform data-level parallelism.

• Proceedings of the Korea Database Society Conference
• /
• 2001.06a
• /
• pp.362-368
• /
• 2001

본 논문에서는 고객의 특성을 고려한 최적의 추천시스템을 개발하기 위하여 기존의 인구통계학적 특성에 따른 협동적필터링 기법의 추천 효율을 비교 분석하였다. 비디오에 대한 사용자 평가 값과 예측 값간의 추천 효율에 대한 비교실험을 통하여 상품에 대한 단순한 선호도만을 고려한 기존의 협동적필터링 방법에 의한 추천시스템의 문제점을 개선하여 추천된 상품이나 콘텐츠에 대한 개인별 추천 효율을 향상시키기 위한 모델을 제시하였다. 본 연구 결과를 이용하여 인터넷 비즈니스 분야에서 활발하게 도입되고 있는 eCRM 시스템에서 가장 중요한 요소인 고객들의 인구통계학적인 다양한 특성을 고려한 협동적필터링 기반의 추천시스템을 개발할 수 있으리라 기대한다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.10a
• /
• pp.268-269
• /
• 2015

The current Internet environment, it is accessible by a computer, but also transferred to the IoT(Internet of Things). These data become large. If the data are provided to the application without any adjustment, it is difficult to exert the original performance. In this paper, we propose a method for filtering the data using the MapReduce of big data processing techniques to refine the collected data. We want to address the heterogeneity of the data generated by the sensor by adding a knowledge identification step in MapReduce. We use XMDR for this purpose.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.2
• /
• pp.119-126
• /
• 2012

Local monitoring is an effective technique in securing data of wireless sensor networks. Existing solutions require high communication cost for detecting false data and this results in a network lifetime being shortened. This paper proposes novel techniques of monitoring based secure data aggregation and filtering false data in wireless sensor networks. The aim is to reduce energy consumption in securing data aggregation. An aggregator and its monitoring node perform data aggregation in a 60o sextant cluster. By checking Message Authentication Codes (MAC), aggregation data will be dropped by a forward aggregator if data aggregated by the aggregator and data monitored by the monitoring node are inconsistent. The simulation shows that the proposed protocol can reduce the amount of average energy consumption about 64% when comparing with the Data Aggregation and Authentication protocol (DAA)[1]. Additionally, the network lifetime of the proposed protocol is 283% longer than that of DAA without any decline in data integrity.

• Proceedings of the IEEK Conference
• /
• 2003.07b
• /
• pp.831-834
• /
• 2003

본 논문에서 기술하고 있는 디블로킹 필터는 ISO/1EC 14496-2 의 디블로킹 필터링 알고리즘[1][2]을 기반으로 한다. 한 개의 레지스터 뱅크를 이용한 효율적인 데이터 스케줄링을 통해 면적과 전력 측면에서 디블로킹 필터를 사용함으로써 생기는 오버헤드를 최소화 시켰으며, CIF 급 영상을 27MHz 동작주파수에서 실시간으로 처리할 수 있도록 설계 하였다. 0.25㎛ Standard Cell Library 로 합성한 결과 총 9800 게이트로 구성 되었으며, 외부 메모리의 도움 없이 동작 시키기 위해 4.4KByte의 버퍼가 사용되었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.746-749
• /
• 2016

주제 기반 크롤링(Topical Crawling)으로 수집된 문서들은 서로 비슷한 단어들을 가지고 있기 때문에 정작 주어진 주제에 적합하지 않은 문서 들을 포함할 수 있다. 이를 해결하기 위해 특정 주제에 해당하는 문서만을 필터링하는 작업이 필요하다. 본 논문은 화재 뉴스 기사에 대한 필터링을 위해 단어 기반 속성과 어울려 화재 뉴스 기사의 특성을 고려한 메타 데이터 속성을 추출하여 이에 특화된 기계학습 메커니즘을 제안하였다. 제안 기법의 F1-측정치는 92.1 %로서, 현재 최고의 성능을 보이는 SVM, 나이브베이즈 알고리즘보다. 2~3% 개선된 것이다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06d
• /
• pp.1-4
• /
• 2008

블룸 필터는 간단하고, 공간 효율적인 자료 구조이다. 블룸 필터는 확률에 근거하여 어떤 데이터 집합을 표현하며, 어떤 데이터가 특정 데이터 집합에 속하는 지를 검사하는 멤버십 쿼리를 지원한다. 이런 멤버십 쿼리는 긍정 오류를 발생시키지만 블룸 필터의 파라미터들을 조정하여 긍정 오류를 최소화할 수 있다. 블룸 필터는 데이터가 공유의 필요성에 의해 전체 시스템에 걸쳐 물리적으로 퍼져있는 분산 시스템과 많은 양의 데이터를 다루기 위해 데이터베이스를 사용하는 시스템 그리고 실시간으로 멤버십 쿼리를 수행해야 하는 시스템 등에서 널리 사용되고 있다. 본 논문에서는 블룸 필터에 대해 알아보고 시스템의 목적에 따라 다양한 형태로 개량된 블룸 필터들에 대해 소개한다.