• Journal of Digital Convergence
• /
• v.15 no.4
• /
• pp.267-273
• /
• 2017

With the development of various devices and communication technologies, the production and distribution of various multimedia contents are increasing exponentially. In order to retrieve multimedia data such as images and videos, an approach different from conventional text-based retrieval is needed. Color and shape are key features used in content-based image retrieval, which quantifies and analyzes various physical features of images and compares them to search for similar images. Color and shape have been used as independent features, but the two features are closely related in terms of cognition. In this paper, a method of describing the spatial distribution of color using a complex color model that projects three-dimensional color information onto two-dimensional complex form is proposed. Experimental results show that the proposed method can efficiently represent the shape of spatial distribution of colors by frequency transforming the complex image and reconstructing it with only a few coefficients in the low frequency.

• Journal of Digital Convergence
• /
• v.17 no.12
• /
• pp.255-260
• /
• 2019

Real-time search keywords are not only limited to search keywords that are rapidly increasing interest in real-time, but also have a limitation that they are difficult to determine the sustainability as there is a difference in ranking between portal sites. Estimating sustainability for real-time search keywords is significant in terms of overcoming these limitations and providing some predictability. In particular, long-term search keywords that last for more than a month are of high value as long-lasting social issues. Therefore, in this paper, we analyze the interest based on the ranking of the real-time search keywords and the duration based on sustained weeks, days and hours of real-time search keywords by each portal site and the integrated portal site, and then estimating sustainability based on high level of interest and duration, and present a method to derive real-time search issues with high long-term sustainability.

• Phonetics and Speech Sciences
• /
• v.7 no.3
• /
• pp.139-149
• /
• 2015

This paper aims to analyze pronunciation variations of loanwords produced by Korean and improve the performance of pronunciation modeling of loanwords in Korean by using syllable-based segmentation and phonological knowledge. The loanword text corpus used for our experiment consists of 14.5k words extracted from the frequently used words in set-top box, music, and point-of-interest (POI) domains. At first, pronunciations of loanwords in Korean are obtained by manual transcriptions, which are used as target pronunciations. The target pronunciations are compared with the standard pronunciation using confusion matrices for analysis of pronunciation variation patterns of loanwords. Based on the confusion matrices, three salient pronunciation variations of loanwords are identified such as tensification of fricative [s] and derounding of rounded vowel [ɥi] and [$w{\varepsilon}$]. In addition, a syllable-based segmentation method considering phonological knowledge is proposed for loanword pronunciation modeling. Performance of the baseline and the proposed method is measured using phone error rate (PER)/word error rate (WER) and F-score at various context spans. Experimental results show that the proposed method outperforms the baseline. We also observe that performance degrades when training and test sets come from different domains, which implies that loanword pronunciations are influenced by data domains. It is noteworthy that pronunciation modeling for loanwords is enhanced by reflecting phonological knowledge. The loanword pronunciation modeling in Korean proposed in this paper can be used for automatic speech recognition of application interface such as navigation systems and set-top boxes and for computer-assisted pronunciation training for Korean learners of English.

• Journal of Convergence for Information Technology
• /
• v.7 no.6
• /
• pp.53-59
• /
• 2017

This study was attempted to identify the domestic research related to depression and stress. The subjects of the analysis were 1,875 college degree theses thrown in the National Assembly Library searched by the depression and stress keyword as of November 30, 2016. The analysis method visualizes atypical data with Word Cloud, which is one of the text mining techniques. We also used the R'LDA package and LDA to classify treatment and subjects. As a result of the analysis, 233(12.4%) of the total papers with therapeutic keywords were found. Application of treatment methods was art therapy, music therapy, horticultural therapy, cognitive behavior therapy, clinical art therapy, cognitive therapy, psychological therapy, depression treatment, group therapy, laughter treatment sequence. The study subjects were adolescents, elderly, patient, mother, child, female, parents, and college students in order. The results of LDA topic analysis for adolescents were classified into four topics: self-support, treatment program, relationship effect, and variable study.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.3 s.35
• /
• pp.55-64
• /
• 2005

The avatar can be used as an auxiliary methodology of text and image communications in cyber space. An intelligent communication method can also be utilized to achieve real-time communication, where intelligently coded data (joint angles for arm gestures and action units for facial emotions) are transmitted instead of real or compressed pictures. In this paper. for supporting the action of arm and leg gestures, a method of generating the facial expressions that can represent sender's emotions is provided. The facial expression can be represented by Action Unit(AU), in this paper we suggest the methodology of finding appropriate AUs in avatar models that have various shape and structure. And, to maximize the efficiency of emotional expressions, a comic-style facial model having only eyebrows, eyes, nose, and mouth is employed. Then generation of facial emotion animation with the parameters is also investigated.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.1154-1158
• /
• 2010

최근 전 세계적으로 GIS (Geographic Information Science) 및 RS (Remote Sensing) 데이터 등 디지털정보의 구축이 급속도로 진행되고 있고, 이들의 발달로 유역에 대한 정확하고 상세한 각종 수문매개변수 수집이 가능하여 유역을 부분유역으로 분할한 기존의 집중형 수문모형보다 유역내의 공간적인 유량변동을 보다 상세하게 고려할 수 있는 격자기반의 분포형 수문모형의 활용도가 높아지고 있다. 유역의 수문특성 및 지형특성을 동일한 매개변수로 적용하기 때문에 유역의 공간적인 수문 및 지형특성을 표현하기 어려운 집중형 모형과 달리 강우-유출해석에 있어서 분포형 모형은 실제 복잡한 유역에서의 유출과정 또는 물질의 수문순환과정을 잘 이해할 수 있고, 어떤 유역의 토지이용형태의 변화가 초래하는 영향과 효과를 사전에 예측할 수 있으며, 신뢰성 있는 과거의 수문자료가 없거나 부족한 유역에서의 유출 계산이 용이하다. 따라서 본 연구에서는 Fortran 90을 개발언어로 사용하여 GIS Data와 위성영상을 활용해 유출량을 모의하는 분포형 물수지 유출 모형을 개발하여 금강 상류유역인 용담댐 유역($930km^2$)을 대항으로 2000~2008년의 일 유출량을 모의하였다. 모형은 크게 3개의 모듈(유출량, 증발산량, 토양수분) 형태로 구성되었으며, 유출량은 강우 전 토양의 저류능을 추적하여 산정하였다. 모형의 결과는 셀별 값을 가지는 분포형으로 출력되며, 유역의 평균 수문자료가 Text file로 출력된다. 민감도 분석을 통하여 최적의 유출 관련 매개 변수를 선정하고 하류의 댐 유입량 자료를 바탕으로 모형의 보정(2001-2004) 및 검증(2005-2008)을 실시하였다. 유출량에 대한 Nash-Sutcliffe 모형효율은 0.78~0.93로 모의치가 실측치의 경향을 잘 표현하는 것으로 나타났다. 유출량 분포도는 강우량을 매우 잘 반영하였으며, 같은 강우조건하에서 토양의 배수조건에 따라 유출이 확연히 다르게 표현되었다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.2
• /
• pp.111-118
• /
• 2015

Entity linking is to link entity's name mentions occurring in text to corresponding entities within knowledge bases. Since the same entity mention may refer to different entities according to their context, entity linking needs to deal with entity disambiguation. Most recent works on entity disambiguation focus on semantic relatedness between entities and attempt to integrate semantic relatedness with entity prior probabilities and term co-occurrence. To the best of my knowledge, however, it is hard to find studies that analyze and present the pure effects of semantic relatedness on entity disambiguation. From the experimentation on Korean Wikipedia data set, this article empirically evaluates entity disambiguation approaches using semantic relatedness in terms of the following aspects: (1) the difference among semantic relatedness measures such as NGD, PMI, Jaccard, Dice, Simpson, (2) the influence of ambiguities in co-occurring entity mentions' set, and (3) the difference between individual and collective disambiguation approaches.

• Journal of Korean Elementary Science Education
• /
• v.24 no.2
• /
• pp.123-129
• /
• 2005

The purpose of this study was to investigate the interaction patterns between teachers and students appeared in sixth grade elementary science lessons. To this end, we set research questions as what are the types of SPS (Social Participation Structure) between teachers and students in the class. Eighteen elementary science lessons were observed. Data were collected by video recordings and observation method. The result showed that the types of SPS were classified into nine. 1) A teacher initiates the whole students into interaction with him. 2) The whole students keep silent or carry out specific activities under the teacher's direction. 3) A teacher initiates some students into interaction. 4) A teacher names specific students to read a text. 5) A teacher initiates the whole students into group interactions. 6) Students interact with each other irrespective of the teacher's intention. 7) Students initiate a teacher into interactions with them. 8) A teacher intervenes during the students' activities and this intervention, in turn, produces interactions between the teacher and the students. 9) A teacher directs the whole class to read a learning goal or a learning material. This study shows that SPS-1 and SPS-3 are the most frequently observed interaction patterns in the elementary science lessons.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.9 no.1
• /
• pp.75-81
• /
• 2016

The main characters play leading roles in novels. In the previous studies, they recognize the main characters in a novel mainly based on dictionaries that built beforehand. In English, names begin with upper cases and are used with some words. In this paper, we propose a recognition method for main characters name in Korean novels by using predicates, rules and weights. We first recognize candidates for the characters name by predicates and propose some rules to exclude candidates that cannot be characters. We assign importances for candidates, considering weights that given by the number of candidates appeared in a sentence. Finally, if the importance of the character is more than a threshold, we decide that the character is one of main characters. The results from the experiments for 300 novels show that an average accuracy is 85.97%. The main characters name may be used to grasp relationships among characters, character's action and tendency.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• v.1
• /
• pp.59-66
• /
• 2006

This research is based on the information compiled on the occurrence of maritime casualties onboard ships registered under Central American Region (CAR) flag States. Due to nonexistence of writing reports in anyone of the countries that are integrating this Region, the information was compiled by the author through personal communication with the people involved in the casualties. The information was compiled by typing up the text and digitizing on the computer for the respective data base. For each casualty occurred, the following information was compiled, date and type of the casualty, position of the ship at the time of the casualty, dimension of the ship, number of people affected by the casualty, and cause and consequences of the casualty. Based on the number of people affected (death and disappear) by the casualties occurred was calculated the mortality rate by 100,000 persons and then compared with the Japanese mortality rate. Furthermore were analyzed the CAR maritime authorities competences and then were compared to that the Japanese counter part. In addition, the implementation rate of the International Maritime Organization (IMO) and International Labor Organization (ILO) conventions ratified by the CAR countries were analyzed. The objective of this research was to compiled and analyze the occurrences of maritime casualties happened on board the ships registered under the CAR Flag States, in order to determine the causes of these accidents. The results of the analysis enable us to better understand of the maritime safety situation of the ships that are registered under the CAR flag States. In order to reduce the occurrence of maritime casualties are proposed a series of measures based on the differences found then the comparison between CAR and Japan. Based on the results of this research, is possible to conclude that the inaccessibility of atmospheric information and the lack of safety measures onboard has been the main cause of the maritime casualties happened in the Pacific side of the CAR.