• Phonetics and Speech Sciences
• /
• v.11 no.3
• /
• pp.39-47
• /
• 2019

We propose a method to improve the performance of spontaneous speech recognizers by extending their phone set using speech data. In the proposed method, we first extract variable-length phoneme-level segments from broadcast speech signals, and convert them to fixed-length latent vectors using an long short-term memory (LSTM) classifier. We then cluster acoustically similar latent vectors and build a new phone set by choosing the number of clusters with the lowest Davies-Bouldin index. We also update the lexicon of the speech recognizer by choosing the pronunciation sequence of each word with the highest conditional probability. In order to analyze the acoustic characteristics of the new phone set, we visualize its spectral patterns and segment duration. Through speech recognition experiments using a larger training data set than our own previous work, we confirm that the new phone set yields better performance than the conventional phoneme-based and grapheme-based units in both spontaneous speech recognition and read speech recognition.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.5
• /
• pp.115-121
• /
• 2019

In this paper, we propose MSAG-SCS-MMA-I adaptive blind equalization with double step size with very small residual ISI and MSE at steady-state while significantly improving the convergence speed of the traditional SCS-MMA-I algorithm in 256-QAM system. And we evaluate the equalization performance for this algorithm. Different step sizes according to the absolute value of decision-directed error instead of a fixed step-size are applied to the tap update equation of MSAG-SCS-MMA-I, which is controlled by binary flags of '1' or '0' obtained from SCS-MMA-I and decision-directed algorithms. This makes for excellent equalization performance. As a result of computer simulation, we confirmed that the proposed algorithm has more better performance than the MMA, SCS-MMA-I, and MSAG-SCS-MMA-I algorithms in terms of the performance index such as residual ISI, MSE, and MD.

• Journal of the Korean Institute of Rural Architecture
• /
• v.24 no.3
• /
• pp.39-46
• /
• 2022

This study aims to derive basic living service facility items in rural areas and construct related spatial data. To do this, a literature review on the laws and systems related to the residential environment and services in rural areas, rural spatial planning, and the 'Rural Convention' strategic plan reports for the Jeolla and Gyeongsang Region in 2021 was conducted. Primary data collection and review on the list of basic living service items in rural areas derived from the analysis were conducted. After data collection, 12 sectors and 44 types of rural basic living service items were derived; the data selection was carried out based on the clarity of the subject of data management, whether it was established nationwide, whether it was disclosed and provided, whether it was periodically updated, and whether it was an underlying law. Afterwards, data on the derived rural basic living service items were constructed. Afterwards, spatial data on the derived rural basic living service items were constructed. Because open data provided through various institutions were employed, data structure unification such as data attribute values and code names was needed, and abnormal data such as address errors and omissions were refined. After that, the data provided in text form was converted into spatial data through geocoding, and through comparative review of the distribution status of the converted data and the provided address, spatial data related to rural basic living services were finally constructed for about 540,000 cases. Finally, implications for data construction for diagnosing rural living areas were derived through the data collection and construction process. The derived implications include data unification, data update system establishment, the establishment of attribute values necessary for rural living area diagnosis and spatial planning, data establishment plan for facilities that provide various services, rural living area analysis method, and diagnostic index development. This study is meaningful in that it laid the foundation for data-based rural area diagnosis and rural planning, by selecting the basic rural living service items, and constructing spatial data on the selected items.

• Journal of KIISE:Databases
• /
• v.30 no.6
• /
• pp.651-665
• /
• 2003

To reduce cache misses emerges as the most important issue in today's situation of main memory databases, in which CPU speeds have been increasing at 60% per year, and memory speeds at 10% per year. Recent researches have demonstrated that cache-conscious index structure such as the CR-tree outperforms the R-tree variants. Its search performance can be poor than the original R-tree, however, since it uses a lossy compression scheme. In this paper, we propose alternatively a cache-conscious version of the R-tree, which we call MR-tree. The MR-tree propagates node splits upward only if one of the internal nodes on the insertion path has empty room. Thus, the internal nodes of the MR-tree are almost 100% full. In case there is no empty room on the insertion path, a newly-created leaf simply becomes a child of the split leaf. The height of the MR-tree increases according to the sequence of inserting objects. Thus, the HeightBalance algorithm is executed when unbalanced heights of child nodes are detected. Additionally, we also propose the CCMR-tree in order to build a more cache-conscious MR-tree. Our experimental and analytical study shows that the two-dimensional MR-tree performs search up to 2.4times faster than the ordinary R-tree while maintaining slightly better update performance and using similar memory space.

• Journal of Internet Computing and Services
• /
• v.19 no.1
• /
• pp.67-76
• /
• 2018

This paper addresses the problem of efficient processing of region monitoring queries. The centralized methods used for existing region monitoring query processing assumes that the mobile object periodically sends location-updates to the server and the server continues to update the query results. However, a large amount of location updates seriously degrade the system performance. Recently, some distributed methods have been proposed for region monitoring query processing. In the distributed methods, the server allocates to all objects i) a resident domain that is a subspace of the workspace, and ii) a number of nearby query regions. All moving objects send location updates to the server only when they leave the resident domain or cross the boundary of the query region. In order to allocate the resident domain to the moving object along with the nearby query region, we use a query index structure that is constructed by splitting the workspace recursively into equal halves. However, However, the above index structure causes unnecessary division, resulting in deterioration of system performance. In this paper, we propose an adaptive split method to reduce unnecessary splitting. The workspace splitting is dynamically allocated i) considering the spatial relationship between the query region and the resultant subspace, and ii) the distribution of the query region. We proposed an enhanced QR-tree with a new splitting method. Through a set of simulations, we verify the efficiency of the proposed split methods.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.6
• /
• pp.1-11
• /
• 2015

Recently, NAND flash memories are used for storage devices because of fast access speed and low-power. However, applications of FTL on low power computing devices lead to heavy workloads which result in a memory requirement and an implementation overhead. Consequently, studies of B+-Tree on embedded devices without the FTL have been proposed. The studies of B+-Tree are optimized for performance of inserting and updating records, considering to disadvantages of the NAND flash memory that it can not support in-place update. However, if a general garbage collection method is applied to the previous studies of B+-Tree, a performance of the B+-Tree is reduced, because it generates a rearrangement of the B+-Tree by changing of page positions on the NAND flash memory. Therefor, we propose a novel garbage collection method which can apply to the B+-Tree based on the NAND flash memory without the FTL. The proposed garbage collection method does not generate a rearrangement of the B+-Tree by using a block information table and a proxy block. We implemented the B+-Tree and ${\mu}$-Tree with the proposed garbage collection on physical devices with the NAND flash memory. In experiment results, the proposed garbage collection scheme compared to greedy algorithm garbage collection scheme increased the number of inserted keys by up to about 73% on B+-Tree and decreased elapsed time of garbage collection by up to about 39% on ${\mu}$-Tree.

• Journal of Korea Spatial Information System Society
• /
• v.6 no.1 s.11
• /
• pp.73-85
• /
• 2004

Recently, as the growth of the wireless Internet, PDA and HPC, the focus of research and development related with GIS(Geographic Information System) has been changed to the Real-Time Mobile GIS to service LBS. To offer LBS efficiently, there must be the Real-Time GIS platform that can deal with dynamic status of moving objects and a location index which can deal with the characteristics of location data. Location data can use the same data type(e.g., point) of GIS, but the management of location data is very different. Therefore, in this paper, we studied the Real-Time Mobile GIS using the HBR-tree to manage mass of location data efficiently. The Real-Time Mobile GIS which is developed in this paper consists of the HBR-tree and the Real-Time GIS Platform HBR-tree. we proposed in this paper, is a combined index type of the R-tree and the spatial hash Although location data are updated frequently, update operations are done within the same hash table in the HBR-tree, so it costs less than other tree-based indexes Since the HBR-tree uses the same search mechanism of the R-tree, it is possible to search location data quickly. The Real-Time GIS platform consists of a Real-Time GIS engine that is extended from a main memory database system. a middleware which can transfer spatial, aspatial data to clients and receive location data from clients, and a mobile client which operates on the mobile devices. Especially, this paper described the performance evaluation conducted with practical tests if the HBR-tree and the Real-Time GIS engine respectively.

• The KIPS Transactions:PartD
• /
• v.9D no.5
• /
• pp.801-812
• /
• 2002

On-line reorganization in a shared nothing database cluster is crucial to the performance of the database system in a dynamic environment like WWW where the number of users grows rapidly and changing access patterns may exhibit high skew. In the existing method of on-line reorganization have a drawback that needs excessive data migrations in case more than two nodes within a cluster have overload at the same time. In this paper, we propose an advanced B$^{+}$ tree based on-line reorganization method that solves data skew on multi-nodes. Our method facilitates fast and efficient data migration by including spare nodes that are added to cluster through on-line scaling. Also we apply CSB$^{+}$ tree (Cache Sensitive B$^{+}$ tree) to our method instead of B$^{+}$ tree for fast select and update queries. We conducted performance study and implemented the method on Ultra Fault-Tolerant Database Cluster developed for high scalability and availability. Empirical results demonstrate that our proposed method is indeed effective and fast than the existing method. method.

• Journal of KIISE:Computer Systems and Theory
• /
• v.34 no.2
• /
• pp.73-83
• /
• 2007

In network intrusion prevention, attack packets are detected and filtered out based on their attack signatures. Pattern matching is extensively used to find attack signatures and the most time-consuming execution part of Network Intrusion Prevention Systems(NIPS). Pattern matching is usually accelerated by hardware and should be performed at wire speed in NIPS. However, that alone is not good enough. First, pattern matching hardware should be able to generate sufficient pattern match information including the pattern index number and the location of the match found at wire speed. Second, it should support pattern grouping to reduce unnecessary pattern matches. Third, it should always have a constant worst-case performance even if the number of patterns is increased. Finally it should be able to update patterns in a few minutes or seconds without stopping its operations, We propose a system architecture to meet the above requirement. The system architecture can process multiple pattern characters in parallel and employs a pipeline architecture to achieve high speed. Using Xilinx FPGA simulation, we show that the new system stales well to achieve a high speed oner 10Gbps and satisfies all of the above requirements.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.1
• /
• pp.77-82
• /
• 2017

This paper related with the VSS_SE-MMA (Varying Step Size_Signed Error-MMA) which possible to improving the equalization performance that employing the varying adaptive step size based on the nonlinearities of error signal of SE-MMA (Signed Error-MMA), compensates the intersymbol interference by distortion occurs at the communication channel, in the transmitting the spectral efficient nonconstant modulus signal such as 16-QAM. The SE-MMA appeared to the reducing the computational arithematic operation using the polarity of error signal in the updating the tap coefficient of present MMA adaptive equalizer, but have a problem of equalization performance degradation. The VSS_SE-MMA improves the problem of such SE-MMA, using the varying step size consider the error signal in the update the adaptive equalizer tap coefficient, and its improved performance were confirmed by simulation. For this, the output signal constellation of equalizer, the residual isi and maximum distortion, MSE and SER were applied. As a result of computer simulation, it was confirmed that the VSS_SE-MMA algorithm has nearly same in convergence speed and has more good performance in every performance index at the steady state.