• Journal of KIISE:Databases
• /
• v.36 no.1
• /
• pp.44-49
• /
• 2009

In recent years, with the development of portable terminals, various searching services on large data have been provided in portable terminals. In order to search large data, most applications for information retrieval use indexes such as B-trees or R-trees. However, only a small portion of the data set is accessed by users, and the access frequencies of each data are not uniform. The existing indexes such as B-trees or R-trees do not consider the properties of the skewed access patterns. And a cache stores the frequently accessed data for fast access in memory. But the size of memory used in the cache is restricted. In this paper, we propose a new index based on disk, called J-tree, which considers user's search patterns. The proposed index is a balanced tree which guarantees uniform searching time on all data. It also supports fast searching time on the frequently accessed data. Our experiments show the effectiveness of our proposed index under various settings.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.12
• /
• pp.1-9
• /
• 2013

Even though the multiple pages TLB are effective in improving the performance, a conventional method with OS support cannot utilize multiple page sizes in user application. Thus, we propose a new multiple-TLB structure supporting multiple page sizes for high performance and low power consumption without any operating system support. The proposed TLB is organised as two parts of a S-TLB(Small TLB) with a small page size and a L-TLB(Large TLB) with a large page size. Both are designed as fully associative bank structures. The S-TLB stores small pages are evicted from the L-TLB, and the L-TLB stores large pages including a small page generated by the CPU. Each one bank module of S-TLB and L-TLB can be selectively accessed base on particular one and two bits of the virtual address generated from CPU, respectively. Energy savings are achieved by reducing the number of entries accessed at a time. Also, this paper proposed the simple 1-bit LRU policy to improve the performance. The proposed LRU policy can present recently referenced block by using an additional one bit of each entry on TLBs. This method can simply select a least recently used page from the L-TLB. According to the simulation results, the proposed TLB can reduce Energy * Delay by about 76%, 57%, and 6% compared with a fully associative TLB, a ARM TLB, and a Dual TLB, respectively.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04a
• /
• pp.590-592
• /
• 2003

최근 XML이 웹 상의 데이터의 표현, 교환, 중재의 표준으로 각광받으면서 이러한 XML 문서를 효과적으로 저장, 접근 및 검색하기 위한 기법에 대한 연구가 많았으나, 기존의 연구들은 하나의 XML 문서를 저장 및 검색의 대상으로 하는 경우가 대부분이였다. 그러나 XML 문서를 데이터의 표현과 교환의 표준으로 이용하는 애플리케이션의 개발이 점차 활성화됨에 따라 저장해야하는 XML 문서의 수가 크게 증가하면서 의미나 구조적으로 많은 유사성을 지니는 XML 문서들을 함께 효율적으로 저장하고 검색하기 위한 기법의 연구가 요구된다. 따라서 본 논문에서는 의미 및 구조적으로 유사성을 가지는 여러 XML 문서들을 통합하는 기법을 제안한다. 제안된 통합 기법은 같은 DTD나 XML Schema를 가지는 경우와 다른 DTD나 XML Schema를 가지는 경우를 모두 고려한다. 또한 특별한 구조적 정보를 가지지 않는 XML 문서의 경우도 다른 DTD나 XML Schema를 가지는 경우와 마찬가지로 처리함으로써 다양한 XML 문서들에 대한 통합이 가능하도록 한다. 이러한 통합 기법은 중복되는 엘리먼트나 애트리뷰트에 대한 저장 공간의 낭비를 최소화한다. 또한 의미적으로 또는 구조적으로 관련성있는 여러 XML 문서의 부분들을 디스크 상의 페이지내에 서로 가까이 저장할 수 있기 때문에 사용자의 일반적인 질의에 대해 효율적이고 빠른 검색 결과를 유도할 수 있고, I/O 횟수를 줄임으로써 그에 따른 오버헤드를 줄일 수 있는 장점이 있다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.7
• /
• pp.1599-1607
• /
• 2015

Recently, researches have been developing efficient techniques for accessing, querying, and managing XML documents which are frequently used in the Internet. In this paper, we propose a parent-child matrix to cluster XML documents efficiently. A parent-child matrix analyzes both the content and structural features of an XML document. Each cell of a parent-child matrix has either the value of a node in an XML tree or the value of a child node, where a parent-child relationship exists in the XML tree. Then, the similarity between two XML documents can be measured by the similarity between two corresponding parent-child matrices. The experiment shows that our proposed method has good performance.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.288-291
• /
• 2008

전세계 인터넷 사용자 수가 약 12억 6천 만명에 이르며 2012년에는 17억명 이상이 인터넷을 사용할 것이라고 추정되는 가운데 웹 데이터베이스 응용에 대한 수요가 비즈니스를 포함하는 모든 분야에서 급속히 증가되고 있다. [1] 그러나 급증하는 수요에 비해 웹 응용의 생성 및 유지·보수는 대부분 수작업에 의존하기 때문에 많은 시간과 노력이 소요되고 있다. 본 논문에서 소개하는 WebAppGen은 ScriptGen과 InteraceGen으로 구성된 웹 응용 자동 생성기로서 웹 응용을 전자동으로 생성해 생산성을 향상시킨다. ScriptGen은 웹 응용에 필요한 폼들과 이 폼들을 통해 이루어지는 질의에 대해 데이터베이스와 연동하여 처리하는 웹 스크립트들을 자동생성하고 InterfaceGen은 ScriptGen을 통해 생성된 폼들을 엮어 접근하기 쉬운 트리 형태의 인터페이스를 제공한다. 상용 웹 응용 생성기들과 달리, WebAppGen은 웹 응용의 전 과정을 자동화 시킴으로써 생산성을 한 단계 더 향상시키고, 생성된 응용은 직관적인 계층구조를 가지는 트리 형태의 인터페이스를 지원한다. WebAppGen이 생성한 응용에서, 한 개의 폼은 관심의 대상인 엔티티 외에 이 엔티티와 직·간접적으로 연관된 모든 엔티티들에 대한 정보를 표현한다. 또한, WebAppGen은 생성될 응용의 기본 틀인 템플릿을 기반으로 하기 때문에 생성된 응용 및 WebAppGen 자체의 유지·보수를 용이하게 한다.

• Journal of KIISE:Databases
• /
• v.35 no.1
• /
• pp.29-43
• /
• 2008

Although RDF ontologies might be expressed in XML tree model, existing methods for protection of XML documents are not suitable for securing RDF ontologies. The graph style and inference feature of RDF demands a new security model development. Driven by this goal, this paper proposes a new query-oriented model for the RDF ontology access control. The proposed model rewrites a user query using a three-layered ontology view. The proposal resolves the problem that the existing approaches should generate inference models depending on inference rules. Accessible ontology concepts and instances which a user can visit are defined as ontology views, and the inference view defined for controling an inference query enables a controlled inference capability for the user. This paper defines the three-layered view and describes algorithms for query rewriting according to the views. An implemented prototype with its system architecture is shown. Finally, the experiment and comparative evaluation result of the proposal and the previous approach is described.

• Journal of KIISE:Information Networking
• /
• v.31 no.1
• /
• pp.37-52
• /
• 2004

Broadcast has been often used to disseminate the frequently requested data efficiently to a large volume of mobile units over a single or multiple channels. Since the mobile units have limited battery power, the minimization of the access time for the broadcast data is an important problem. There have been many research efforts that focus on the improvement if the broadcast techniques by providing indexes on the broadcast data. In this paper, we studied an efficient index allocation method for the broadcast data over multiple physical channels, which cannot be coalesced into a single high bandwidth channel. Previously proposed index allocation techniques either require the equal size of index and data or have a performance degradation problem when the number of given physical channels is not enough. These two problems will result in the increased average access time for the broadcast data. To cope with these problems, we propose an efficient tree- structured index allocation method for the broadcast data with different access frequencies over multiple physical channels. Our method minimizes the average access time for the broadcast data by broadcasting the hot data and their indexes more often than the less hot data and their indexes. We present an in-0e0th experimental and theoretical analysis of our method by comparing it with other similar techniques. Our performance analysis shows that it significantly decrease the average access time for the broadcast data over existing methods.

• The Journal of the Korea Contents Association
• /
• v.4 no.4
• /
• pp.141-152
• /
• 2004

In this paper, we propose the index structure supporting the future position retrievals with efficiently updating continuous positions of moving objects in location based services. For reducing update costs of moving objects, our index structure directly accesses to the leaf node with moving objects using secondary index structure and performs bottom up update when node information is changed. Positions of moving objects are stored in primary index structure. In primary index structure, the split information similar to kd-tree is stored to internal node for increasing node's fanout. And the proposed index structure supports the future position retrievals using velocity of moving objects in the child node.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4B
• /
• pp.403-418
• /
• 2009

IP version 6 (IPv6) is a new If addressing scheme that has 128-bit address space. IPv6 is proposed to solve the address space problem of IP version 4 (IPv4) which has 32-bit address space. For a given IPv6 routing set, if a forwarding table is built using a trio structure, the trio has a lot more levels than that for IPv4. Hence, for IPv6 address lookup, the binary search on trio levels would be more appropriate and give better search performance than linear search on trio levels. This paper proposes a new IPv6 address lookup algorithm performing binary search on trio levels. The proposed algorithm uses a Bloom filter in pre-filtering levels which do not have matching nodes, and hence it reduces the number of off-chip memory accesses. Simulation has been performed using actual IPv6 routing sets, and the result shows that an IPv6 address lookup can be performed with 1-3 memory accesses in average for a routing data set with 1096 prefixes.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.42 no.10 s.340
• /
• pp.25-30
• /
• 2005

This paper generalizes the concept of the signal space detection to construct a fixed delay tree search (FDTS) detector which estimates a block of n channel symbols at a time. This technique is applicable to high speed implementation. Two approaches are discussed both of which are based on efficient signal space partitioning. In the first approach, symbol detection is performed based on a multi-class partitioning of the signal space. This approach is a generalization of binary symbol detection based on a two-class pattern classification. In the second approach, binary signal detection is combined with a look-ahead technique, resulting in a highly parallel detector architecture.