• The KIPS Transactions:PartD
• /
• v.14D no.7
• /
• pp.719-726
• /
• 2007

Recently, NAND flash memory has been used for a storage device in various mobile computing devices such as MP3 players, mobile phones and laptops because of its shock-resistant, low-power consumption, and none-volatile properties. However, due to the very distinct characteristics of flash memory, disk based systems and applications may result in severe performance degradation when directly adopting them on flash memory storage systems. Especially, when a B-tree is constructed, intensive overwrite operations may be caused by record inserting, deleting, and its reorganizing, This could result in severe performance degradation on NAND flash memory. In this paper, we propose an efficient buffer management scheme, called IBSF, which eliminates redundant index units in the index buffer and then delays the time that the index buffer is filled up. Consequently, IBSF significantly reduces the number of write operations to a flash memory when constructing a B-tree. We also show that IBSF yields a better performance on a flash memory by comparing it to the related technique called BFTL through various experiments.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.11a
• /
• pp.329-332
• /
• 2006

DBMS상의 인덱싱 기법은 대부분 B-트리 또는 B-트리의 변형인 T-트리 등을 주로 사용하여 왔다. 이는 트리 구성에서 최적의 균형을 유지하여 자료검색 및 저장 공간에 효율적인 방안으로 인식되어 왔으며 대부분의 범용 DBMS에서 탁월한 성능을 보여 왔다. 그러나 고성능의 효율을 요구하는 OLTP 분야 중 특히 증권시장에서는 이런 일반적인 인덱싱 방법보다는 그 분야 특성에 맞는 새로운 인덱싱 방안의 적용이 성능 향상과 더불어 시스템의 최적의 효율을 얻어낼 수 있다. 본 논문에서는 증권시장용 원장(ledger) DB의 새로운 인덱싱 방안으로서, 먼저 원장 DB의 접근 형태를 이해하고 경험적인 데이터를 기준으로 산출한 가중치를 이용, 새로운 인덱싱 방안을 제시하고 기존 인덱싱 방법과의 비교를 통한 성능향상의 효율성을 보여준다.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.10
• /
• pp.2615-2628
• /
• 1997

In this paper, a new codebook design algorithm is proposed. It uses a DCT map based on two-dimensional discrete cosine of transform (2D DCT) and finite state vector quantizer (FSVQ) when the vector quantizer is designed for image transmission. We make the map by dividing input image according to edge quantity, then by the map, the significant features of training image are extracted by using the 2D DCT. A master codebook of FSVQ is generated by partitioning the training set using binary tree based on tree-structure. The state codebook is constructed from the master codebook, and then the index of input image is searched at not master codebook but state codebook. And, because the coding of index is important part for high speed digital transmission, it converts fixed length codes to variable length codes in terms of entropy coding rule. The huffman coding assigns transmission codes to codes of codebook. This paper proposes single-side growing huffman tree to speed up huffman code generation process of huffman tree. Compared with the pairwise nearest neighbor (PNN) and classified VQ (CVQ) algorithm, about Einstein and Bridge image, the new algorithm shows better picture quality with 2.04 dB and 2.48 dB differences as to PNN, 1.75 dB and 0.99 dB differences as to CVQ respectively.

• The Journal of Society for e-Business Studies
• /
• v.9 no.1
• /
• pp.83-104
• /
• 2004

Recently, information systems that require storage and retrieval of huge amount of data are becoming used widely. Accordingly, research efforts have been made to develop Linux cluster file systems in the SAN environment in which clients themselves can manage metadata and access data directly. Also a semi-flat directory structure based on extendible hashing has been proposed to support fast retrieval of files[1]. In this research, we have designed and implemented the semi-flat extendible hash directory under the Linux system. In order to evaluate the practicality of the directory, we have also implemented the B+-tree based directory and experimented the performance. According to the performance comparisons, the extendible hash directory has the better performance at insert, delete, and search operations. On the other hand, the B+-tree directory is better at sorting files.

• The KIPS Transactions:PartB
• /
• v.10B no.5
• /
• pp.575-578
• /
• 2003

SPIHT algorithm is a wavelet based fast and effective technique for image compression. It uses a list structure to store status information which is generated during set-partitioning of zero-tree. Usually, this requires lots of additional memory depending on how high the bit-rate is. Therefore, in this paper, we propose a new technique called MZP-DFS, which needs no additional memory when running SPIHT algorithm. It traverses a spatial-tree according to DFS and eliminates additional memory as it uses test-functions for encoding and LSB bits of coefficients for decoding respectively. This method yields nearly the same performance as SPIHT. This may be desirable in hardware implementation because no additional memory is required. Moreover. it exploits parallelism to process each spatial-tree that it can be applied well in real-time image compression.

• The KIPS Transactions:PartB
• /
• v.12B no.2 s.98
• /
• pp.215-222
• /
• 2005

In this paper, a MetaSearch algorithm for navigational queries is presented. Previous MetaSearch algorithms focused on informational queries. They Eave a high score to an overlapped document. However, the overemphasis of overlapped documents may degrade the performance of a MetaSearch algerian for a navigational query. However, if a lot of result documents are from a certain domain or a directory, then we can assume the importance of the domain or directory. Various experiments are conducted to show the effectiveness of overlap of a domain and directory names. System results from TREC and commercial search engines are used for experiments. From the results of experiments, the overlap of documents showed the better performance for informational queries. However, the overlap of domain names and directory names showed the $10\%$ higher performance for navigational queries.

• Journal of Convergence for Information Technology
• /
• v.11 no.3
• /
• pp.212-219
• /
• 2021

The current study aimed to find the effects of hair treatment with shea butter on bleached hair and hair protection. The agent for hair treatment with shea butter was developed in different concentration levels (0%,1%,3%,5%) for experimentation. After applying the agent to bleached hair, hair samples were evaluated as follows. Increase in hair thickness was highest in the hair sample that had 5% of shea butter. The amount of amino acids was also highest in the hair sample that contained 5% of shea butter. Comparing the difference of the surface color, hair sample with 5% of shea butter showed low level of L⁎ while the level of a⁎ value that reflects the redness was high and the level of b⁎ value that reflects yellowness was low. Observation through the scanning microscope confirmed the positive effects by showing smoother surfaces in the sample with shea butter than in the bleached hair. This study showed shea butter is considered to be cosmetics to protect damaged hair by bleaching.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.1
• /
• pp.187-194
• /
• 1996

In this paper, we propose regression neural networks based on regression trees. We map regression trees into three layered feedforward networks. We put multi feature split functions in the first layer so that the networks have a better chance to get optimal partitions of input space. We suggest two supervised learning algorithms for the network training and test both in single feature split and multifeature split functions. In experiments, the proposed regression neural networks is proved to have the better learning performance than those of the single feature split regression trees and the single feature split regression networks. Furthermore, we shows that the proposed learning schemes have an effect to prune an over-grown tree without degrading the learning performance.

• The KIPS Transactions:PartD
• /
• v.12D no.6 s.102
• /
• pp.915-920
• /
• 2005

In this paper we expanded a existing web based B2B system to support users in low speed network. To guarantee shared dat a consistency between clients and a server, we proposed a method of data change detection by using a time stamp tree and the performance analysis of the proposed method was proved by a simulation. Under the worst condition that leaf nodes of a times tamp tree were changed uniform distribution, the simulation result showed that the proposed method was more efficient than a sequential detection until the percentage of changed nodes were below $15\%$. According to our observation, the monthly average of data change was below $7\%$ on a web-based construction MRO B2B system or a company A from April 2004 to August 2004. Therefore the Proposed method improved performance of data change detection in practice. The proposed method also reduced storage consumption in a server because it didn't require a server to store replicated data for every client.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04b
• /
• pp.67-69
• /
• 2001

최근 이미지나 멀티미디어 데이터와 같은 비정형 데이터의 검색을 보다 효과적으로 수행하기 위한 연구가 활발하게 진행되어 왔다. 비정형 데이터를 검색하기 위해서는 비정형 데이터를 다차원의 특징 벡터로 변환하고, 그것을 다차원 색인구조를 이용해 색인한다. 따라서 이러한 비정형 데이터를 효율적으로 색인 할 수 있는 다차원 색인구조가 요구되고 있다. 이 논문에서는 데이터를 벡터 근사치로 표현한 후 이를 트리 형태로 구성하여 검색이 효율을 높이는 다차원 데이터를 위한 색인구조 VA(Vector Approximate)-트리를 제안한다. 이 논문에서 제안하는 VA-트리는 VA-파일과 K-D-B-트리 구조를 기반으로 하고 있다. VA-트리는 적은 비트를 이용하여 다차원 공간을 표현하기 위해 노드내의 모든 정보를 비트로 표현한다. 중간노드의 비트 형태 엔트리는 하위노드에 포함된 정보를 의미하고 있어 탐색을 효율적으로 수행할 수 있도록 한다. 실험을 통한 성능평가를 수행하여 제안된 색인구조의 우수함을 보인다.