• Journal of Korean Institute of Industrial Engineers
• /
• v.39 no.2
• /
• pp.119-128
• /
• 2013

Most internet users utilize internet portal search engines, such as Naver, Daum and Google nowadays. But since the results of internet portal search engines are based on universal criteria (e.g. search frequency by region or country), they do not consider personal interests. Namely, current search engines do not provide exact search results for homonym or polysemy because they try to serve universal users. In order to solve this problem, this research determines keyword importance and weight value for each individual search characteristics by collecting and analyzing customized keyword at external database. The customized keyword weight values are integrated with search engine results (e.g. PageRank), and the search ranks are rearranged. Using 50 web pages of Goolge search results for experiment and 6 web pages for customized keyword collection, the new customized search results are proved to be 90% match. Our personalization approach is not the way that users enter preference directly, but the way that system automatically collects and analyzes personal information and then reflects them for customized search results.

• The Journal of the Korea Contents Association
• /
• v.7 no.2
• /
• pp.73-83
• /
• 2007

An information retrieval system, using exact match, needs preprocessing or query expansion to generate transliteration variants in order to search foreign word transliteration variants in the documents. This paper proposes an effective method to generate other transliteration variants from a given transliteration. Because simple rewriting of confused characters produces too many false variants, the proposed method controls the generation priority by learning confusion patterns from real uses and calculating their probability. Especially, the left and right context of a pattern is considered, and local rewriting probability and global rewriting probability are calculated to produce more probable variants in earlier stage. The experimental result showed that the method was very effective by showing more than 80% recall with top 20 generations for a transliteration variants set collected from KT SET 2.0.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.21 no.8
• /
• pp.151-159
• /
• 2007

This paper proposes a improved decoupling current control algorithms using a compensator to enhance the robustness to the stator inductance error. And the compensator composed from a d-axis current error for sensing change of the real stator inductance can match the estated stator inductance to the real stator inductance. A great advantage of this algorithms is more robust than feedforward decoupling current control or dynamic decoupling control. So it looks suitable to the practical applications where the exact parameters are unknown. Though simulation and experimental results with 1[kW] prototype PMSM and TMS320F2812 board for motor control, we show that the proposed controller achieves the desired performance.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.23 no.11 s.170
• /
• pp.1940-1951
• /
• 1999

In this paper, the equations of motion of a Bernoulli-Euler cantilever beam fixed on a moving cart and carrying a lumped mass concentrated at an arbitrary position along the beam is derived. The motion of the beam-mass-cart system is analyzed through unconstrained modal analysis, and a unified characteristic equation for calculating the natural frequencies of the system is obtained. The changes of natural frequencies and the corresponding mode shapes with respect to the changes in mass ratios of the system and to the concentrated position of the lumped mass are investigated with the frequency equation, which can be generally applied to this kind of systems. The exact and assumed-mode solutions including the dynamics of the base cart are obtained, and the open-loop responses of the system by arbitrarily designed forcing function are given by numerical simulations. The results match well with physical phenomena even at the extreme cases where the concentrated mass is attached to the bottom and to the top of the beam.

• Journal of Internet Computing and Services
• /
• v.1 no.1
• /
• pp.1-14
• /
• 2000

Visual information retrieval and image databases are very important applications of spatial access methods. The quaries for these applications are visual and based not on exact match but on dubjective similarity. The individual aperations of spatial access methods are much more expensive than those of conventional one-dimensional access methods. Also, because the visual queries are much more complex than textual queries, an efficient processing technique for visual queries is one of the critical requirements in the development of large and scalable image databases. Therefore, efficient translation and execution for the complex visual queries are not less important than those of textual databases. In this paper, we introduce our cognitive and topological studies that are required to process subjective visual queries effectively. Then, we propose an efficient translation and execution techniques for similarity based visual queries by conducting these related studies.

• ETRI Journal
• /
• v.41 no.6
• /
• pp.771-781
• /
• 2019

We propose the VS³-NET model to solve the task of question answering questions with machine-reading comprehension that searches for an appropriate answer in a given context. VS³-NET is a model that trains latent variables for each question using variational inferences based on a model of a simple recurrent unit-based sentences and self-matching networks. The types of questions vary, and the answers depend on the type of question. To perform efficient inference and learning, we introduce neural question-type models to approximate the prior and posterior distributions of the latent variables, and we use these approximated distributions to optimize a reparameterized variational lower bound. The context given in machine-reading comprehension usually comprises several sentences, leading to performance degradation caused by context length. Therefore, we model a hierarchical structure using sentence encoding, in which as the context becomes longer, the performance degrades. Experimental results show that the proposed VS³-NET model has an exact-match score of 76.8% and an F1 score of 84.5% on the SQuAD test set.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.22 no.7
• /
• pp.40-47
• /
• 2014

This paper proposes a matching method of GIS lane information with a vehicle position using camera image to reduce DGPS error. Images of straight road are taken using a camera that is installed on the front center of the vehicle, and the distance between the vehicle and the lane are estimated using the images. The current GIS lane data is matched by comparing the estimated distance and the measured distance using a DGPS. Inverse perspective mapping is used to minimize the error of image processing from the heading angle, and single buffering method is applied to decide the exact moment of GIS match. Through practical test on the highway, feasibility of the GIS matching using camera image is confirmed.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.13 no.5
• /
• pp.875-885
• /
• 2010

Recently, the role of Ubiquitous Sensor Network(USN) has been considered to be essential for supporting the near future Network Centric Warfare(NCW) and Tactical Information Communication Network(TICN). In this paper, we explore a set of data storage methods(external storage, local storage and data storage) and query processing methods in WSN. In particular, we focus on analyzing a novel data structure for supporting the local storage method, named the partial ordered tree(POT). The main idea behind POT is that sensor readings are usually correlated with the physical spatial domain. With the help of POT, only a small portion of sensor nodes participate in query processing tasks, and thus network lifetime is greatly increased. Through a series of simulation experiments, we demonstrate that the POT based local storage method clearly outperforms the existing data storage methods in terms of the energy-efficiency, which directly affects the network lifetime, for processing exact match queries, range queries and top-k queries.

• Journal of Korea Multimedia Society
• /
• v.25 no.11
• /
• pp.1643-1652
• /
• 2022

Most of the existing machine reading research has used Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) algorithms as networks. Among them, RNN was slow in training, and Question Answering Network (QANet) was announced to improve training speed. QANet is a model composed of CNN and self-attention. CNN extracts semantic and syntactic information well from the local corpus, but there is a limit to extracting the corresponding information from the global corpus. Graph Convolutional Networks (GCN) extracts semantic and syntactic information relatively well from the global corpus. In this paper, to take advantage of this strength of GCN, we propose I-QANet, which changed the CNN of QANet to GCN. The proposed model performed 1.2 times faster than the baseline in the Stanford Question Answering Dataset (SQuAD) dataset and showed 0.2% higher performance in Exact Match (EM) and 0.7% higher in F1. Furthermore, in the Korean Question Answering Dataset (KorQuAD) dataset consisting only of Korean, the learning time was 1.1 times faster than the baseline, and the EM and F1 performance were also 0.9% and 0.7% higher, respectively.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.7
• /
• pp.135-142
• /
• 2020

Basketball is a highly complex sport, analyses offensive and defensive rebounds, free throw percentages, minutes played and an efficiency rating. These statistics can have a large bearing and provide a lot of pressure on players as their every move can be analysed. Performance analysis in sport is a vital way of being able to track a team or individuals performance and more commonly used resource for player and team development. Discovering information such as this proves the importance of these types of analysis as with post competition video analysis a coach can reach a far more accurate analysis of the game leading to the ability to coach and correct the exact requirements of the team instead of their perceptions. A significant difference was found between winning and losing performances for different types of turnovers supporting current research that states that turnovers are not a valid predictor of match outcomes and that there is no specific type of turnover which can predict the outcome of a match as briefly mentioned in Curz and Tavares (1998). Significant differences were found between winning and tied and losing and tied performance for some types of turnovers, however due to the lack of data collected in this area they cannot be considered valid. Further research could also be conducted in other areas relating to performance indicators where there is currently minimal research in some areas such as assisted baskets, stated about the performance indicators in their own study the performance indicators are inadequate for explaining the complexities of the game suggesting that one indicator will not be constant in every game an research into performance analysis areas would be more appropriate.