Earticle

Beoncheon-ri white porcelain kiln site located on the territory of Gwangju-si(Gyeonggi-do) was a state porcelain kiln. By epitaph found in the site, it’s dated from the second half of 16th century. There were produced the best products of that time. On the site 2 kilns, 7 workshops, 5 dust holes for pottery rejectsand other objects forming a single production complex were studied. In the article we make morphological analysis of white porcelain bowls found in largest quantities to study the bowls and interrelation between different objects, such as kilns, workshops and dust holes. The objects of analysis were divided in three groups: dust holes 1-2 and 1-4; dust holes 2, 4 and 5 and workshop 4; workshops 1, 2 and 5. Is discovered that these three groups have also relative chronological significance. Judging by stratigraphy and morphological analysis, the first group existed in the earliest period, and the third one in the latest. Besides, it turned out, that all the bowls in a separate group have equal morpheme of an attribute. In general, in the article are offered criteria of the impartial determination of the bowls size and of the distinguishing of rice bowls (bal’) and soup bowls (daejop) by bowl’s height and rim diameter coefficient. The mutual relation between different objects were cleared up, the characteristic morpheme of bowls attributes in every group were analyzed.

본고는 국어사 말뭉치의 구축 현황을 분석하고 국립국어원의 ‘국어 역사 말뭉치’를 대상으로 하여 딥러닝 시스템으로 형태 분석을 시도한 것이다. 현대 국어 연구와 달리 국어사 분야에서는 다양한 디지털 인문학적 방법론 적용에 한계가 있었다. 딥러닝 기반의 국어사 형태 분석을 위해 필요한 세종 말뭉치의 구축 현황을 분석하여 언해문과 원문을 추출하고 각 세기별 국어의 형태론적 특징을 반영할 필요성을 확인하였다. 정밀한 품사 태깅을 위해 학습 데이터의 확장과 모델 고도화가 요구되나 국어사 연구에 딥러닝 기술을 적용하여 형태 분석의 가능성을 모색한 시도로 세기별 형태 분석 모델의 개발과 국어사 정보의 활용을 위해 딥러닝을 활용한 초기 연구로 의의를 지닌다.

This study investigates methods for applying deep learning to the analysis of historical Korean corpus. While contemporary Korean linguistics has widely adopted computational and digital humanities approaches, research on historical Korean has been limited by the complexity of older language forms and the scarcity of annotated data. Focusing on the National Institute of Korean Language’s Historical Korean Corpus, this research examines the construction status of the Sejong Corpus, extracts vernacular translations and original texts, and identifies the need to incorporate century-specific morphological characteristics. The study demonstrates that deep learning-based morphological analysis is feasible but requires both the expansion of training data and the refinement of model architectures to achieve accurate part-of-speech tagging. As one of the earliest attempts to integrate deep learning into historical Korean linguistics, this work highlights the potential for developing century-specific morphological models and for advancing the broader utilization of linguistic information in historical Korean corpus.

본 연구는 트위터에서 추출한 저출산 관련 메시지의 의미연결망 네트워크 구조를 분석하여 저출산 관련 정책 수립에 시사점을 주고자 하였다. 2015년 10월 16일부터 2017년 11월 15일까지 트위터에 등록된 11,984건의 저출산 관련 메시지를 대상으로 형태소 분석을 통해 상위 빈출 단어 110개를 선정하였다. 그리고 상위 110개의 단어를 대상으로 의미연결망 분석을 실시하였다. 분석 결과, 현재의 저출산 상황이 개인의 자발적 선택의 결과라는 것을 알 수 있었다. 반면, 출산 장려 정책과 관련한 단어의 언급은 상당히 미진한 것으로 나타났다. 콘커 분석을 통해 각 단어 간의 의미 연결 구조를 살펴본 결과, 상위 110개의 단어는 총 8개의 그룹으로 구분되었다. 각 그룹별 단어는 저출산 현황에 대한 자신의 주변 상황 및 인식, 가치관의 변화, 결혼과 임신, 출산에 이르는 다양한 걱정이 주류를 이루고 있었다. 이러한 결과를 바탕으로 출산장려 정책에 대한 제언을 담았다.

This study aims to analyze the communication network structure of messages related to low fertility on Twitter to contribute to policy-making of the government. This study selected top 110 words from 11,984 messages, which are posted on Twitter from 16 October 2015 to 17 November 2017, based on the result of the morphological analysis. And these top words were investigated by semantic network analysis. According to the result of these examinations, the majority of tweets point out that on own’s choice is the main reason for low fertility. Meanwhile, the quantity of tweets about childbirth promoting policies is significantly small. In addition, as the result of the concor analysis, top 110 words are divided into eight groups. Respective groups of words represent various issues lead to low fertility, such as negative perceptions toward the environment, changed world view, avoid getting married, and concerns about pregnancy and childbirth. Therefore, this study suggests some practical implications to promote childbirth policies based on these results.

본고는 현대 한국어의 복합 조사와 관련한 문제를 검토하여 복합 조사의 개념을 정립하는 데 연구 목적을 두 었다. 현대 한국어에는 둘 이상의 조사가 선행어에 이어서 쓰인 경우가 많이 있다. 많은 학자들은 이러한 구성을 복 합 조사라고 부른다. 하지만 공시적으로 조사와 조사의 결합은 조사와 결합하는 선행어가 실사이어야 한다는 조건에 위배된다는 문제가 있다. 본고는 이와 같은 문제점을 기술하는 것에서 시작하여, 형태, 통사, 의미적으 로 복합 조사의 특성을 밝히고, 나아가 복합 조사의 개념을 정립하고자 하였다.

This article defines Korean compound particle through studying some related questions on modern Korean compound particle. The structure that an antecedent following with more than two particles in sequence among modern Korean is called compound particle by most scholars. Obviously, the combining of particles goes against with the term, which was determined that the combination with a particle should be an from the perspective of Synchronic. So from the very beginning with the discussion of those questions, the article will illustrate the features of compound particle from different aspects, like morphology, syntax and semantics, etc. Based on it, an definition on compound particle will be concluded.

Morphological analysis is used for searching sentences and understanding context. As most morpheme analysis methods are based on predefined dictionaries, the problem of a target word not being registered in the given morpheme dictionary, the so-called unregistered word problem, can be a major cause of reduced performance. The current practical solution of such unregistered word problem is to add them by hand-write into the given dictionary. This method is a limitation that restricts the scalability and expandability of dictionaries. In order to overcome this limitation, we propose a novel method to automatically expand a dictionary by means of use-case analysis, which checks the validity of the unregistered word by exploring the use-cases through web crawling. The results show that the proposed method is a feasible one in terms of the accuracy of the validation process, the expandability of the dictionary and, after registration, the fast extraction time of morphemes.

Climate change is accelerating shifts in the natural environment, increasing both the frequency of extreme events such as typhoons and floods and the severity of river damage. While natural rivers tend to form meanders that reduce the energy expenditure of flowing water, regulated rivers with hydraulic structures such as levees and weirs strongly constrain this behavior. As a result, the risk of failure at river bends can rise rapidly during floods that exceed the design return period. Because meandering remains a representative geomorphic feature in both natural and regulated rivers, understanding the problems and site-specific phenomena that occur in bends is a key issue in river engineering. This study examines Korean rivers to identify the installation locations of transverse structures (weirs) and their offset distances from river bends, and to characterize these patterns in relation to geomorphic factors and design flood discharge. The results provide baseline information for subsequent mobile-bed experiments in a 90° meandering channel to quantify how weir location affects longitudinal and transverse bed morphology and water-surface superelevation, and to derive bend-design implications under climate-change-driven flood risk.

This study was conducted to estimate climax index by eco-morphology for major 36 tree and sub-tree species in natural deciduous forests so as to interpret seral position of each species in the forest community. Fourteen eco-morphological characters which were considered to be associated with successional gradient in the forest were selected for the study. Four levels per character for each species were given on a standardized scale of increasing climax, and the index was computed by the proportion of the sum of total scores, expressed by percent values. With calculated mean value of 54.8 for all indices, Carpinus cordata had the highest index value of 90.5, and Populus davidiana recorded the lowest of 13.2. The most climax group, greater than 70 of the index, contained only 8 species, intermediate group, between 41 to 70 of the index, had 23 species, and the most pioneer group, less than 40 of the index comprised 5 species. The result has noticed that the large number of species would take advantage of most diverse resource and niche in the intermediate stage of the sere in the forest. By cluster analysis all 36 species were subjected to be classified into several species groups which had common similar eco-morphological characteristics. The indices were additionally plotted on the two dimensional graph to recognize the positions related to the light absorption factor and reproduction factor. The climax index of tree and sub-tree species developed by this study could be applied to understand the present status of successional stage on the basis of species composition by the method of summing up the indices. And comparison of forest successional stage among various forest communities could be done by summing up the climax indices of composed species in each community. However, this kind of applied methodology should be limited to the forest of similar species composition and site condition.

Purpose: This study aimed to generate virtual mandibular first molars using deep learning technology, specifically deep convolutional generative adversarial network (DCGAN), and evaluate the accuracy and reliability of these virtual teeth by analyzing their morphological characteristics. These morphological characteristics were classified based on various evaluation criteria, facilitating the assessment of deep learning-based dental prosthesis production’s practical applicability. Methods: Based on our previous research, 1,000 virtual mandibular first molars were generated, and based on morphological criteria, categorized as matching, non-matching, and partially matching. The generated first molars or the categorization of the generated molars were analyzed through the expert judgment of dental technicians. Results: Among the 1,000 generated virtual teeth, 143 (14.3%) met all five evaluation criteria, whereas 76 (7.6%) were judged as completely non-matching. The most frequent issue, with 781 (78.1%) instances, including some overlapping instances, was related to occlusal buccal cusp discrepancies. Conclusion: The study reveals the potential of DCGAN-generated virtual teeth as substitutes for real teeth; however, additional research and improvements in data quality are necessary to enhance accuracy. Continued data collection and refinement of generation methods can maximize the practicality and utility of deep learning-based dental prosthesis production.

본 연구는 한국어 학습자가 속담을 사용할 때 형태적 측면에서 어떤 양상을 보이는지 살펴본 연구이다. 한국어 학습자 말뭉치에서 속담을 사용한 문장을 추출하여 분석한 결과 속담을 등재형과 활용형으로 사용하는 비율이 높게 나타났다. 이를 통해 한국어 학습자들은 속담을 거의 변형하지 않고 사용하고 있음을 알 수 있다. 속담 말뭉치에 나타난 오류를 살펴보니 문법적 오류의 비중이 압도적으로 높게 나타났다. 학습자들이 속담의 의미나 어휘 등의 이해에 어려움을 겪는다기보다는 인용 표지나 비유 표지 등 속담 사용 시 형태적인 문법을 잘못 쓰는 경우가 많다는 것을 알 수 있다. 그간 속담 교육이 의미 이해 중심으로 이루어져 왔으나 실제 학습자의 사용 양상을 살펴본 결과 속담 사용 시 인용 표지 및 형태와 관련된 교육과 변형형에 대한 교육이 필요함이 드러났다.

This study examines morphological aspects of the use of proverbs by Korean language learners through a corpus analysis of language learners’ sentences containing proverbs. The sentences showed a high proportion of basic forms and inflectional forms, indicating that Korean language learners use proverbs without variation. Moreover, it was found that grammatical errors made up an overwhelming proportion of all errors, showing that learners often make mistakes in the appropriate grammatical forms when using proverbs, such as the quotative verb suffixes or metaphorical signs, rather than having difficulty understanding the meaning or vocabulary of the proverb. While previous proverb education has focused on meaning, this study highlights the need for education based on the usage patterns of learners regarding the expression of quotations, grammatical forms, and variability when using proverbs.

국어에는 ‘하루, 이틀, 사흘, 나흘, 열흘’과 같이 날짜를 지칭하는 어휘가 존재한다. 이들은 중세국어로는 ‘, 이틀, 사/사, 나, 열흘’의 형태를 취하는데, 자세히 분석해 보면 모두 공통적으로 ‘/흘’이 결합되어 있음을 알 수 있다. ‘/흘’은 형태가 ‘/을’일 가능성도 있지만, ‘이틀, 읻, 이듬, 이듭’에서 분석되는 ‘읻’이 ㅎ곡용어와는 다른 특성을 보이기 때문에 선행 구성 성분을 ㅎ곡용어로 보기보다는 후행 구성 요소의 형태를 ‘/흘’로 보는 것이 합리적이다. 이 ‘/흘’은 기존 연구에서 접미사로 분석되기도 했지만, ‘오’(오-+-ㄴ#)에서 관형사형의 수식을 받는 것으로 보아 원래 명사였던 것으로 판단된다. ‘’에는 외견상 ‘/흘’이 결합되어 있지 않은 것처럼 보이지만, 형태소를 분석해 보면 ‘/흘’이 포함되어 있음을 알 수 있으며, 현대국어의 ‘며칠’은 ‘몇’과 ‘일’(日)이 결합한 것이 아니라 ‘몇’과 ‘흘’이 결합한 후 ‘며츨’을 거쳐 ‘며칠’이 된 것으로 분석된다.

Specially we have some words about date. Among them I treated several words in relation to ‘hʌl/hɨl’ which means ‘day’. The subjects of my study are concretely ‘haru, ithɨl, sahɨl, nahɨl, jəlhɨl’. They were spelled ‘hʌrʌ, ithɨl, sahʌl/saʌl, naʌl, jəlhɨl’ in middle Korean. They have something in common in respect of containing of ‘hʌl/hɨl’. The underlying form of this noun is not ‘ʌl/ɨl’ but ‘hʌl/hɨl’. Some linguists asserted that ‘ʌl/ɨl’ was the right underlying form. But ‘ithɨl’ is analyzed as compound of ‘it’ and ‘hɨl’. So I cannot but regard ‘hʌl/hɨl’ as right underlying form. ‘hʌl/hɨl’ is not a suffix but a noun. Because it is modified by prenoun form as ‘onʌl’(o-+-n#hʌl). ‘hʌrʌ’ and ‘mʌrʌ’ is analyzed as respectively ‘*hʌtʌn+hʌl’ and ‘mʌt+ʌl’. And ‘mjəchil’ is analyzed as compound of ‘mjəch’ and ‘hɨl’.

This study utilized scanning electron microscopy (SEM) to analyze the morphological characteristics of hair cut by various methods. In forensic investigations, hair serves as crucial evidence. When the hair root is present, DNA analysis can be employed for individual identification. However, when hair is cut in the middle, DNA analysis becomes challenging, making the morphological analysis of cut marks an essential aspect of forensic examinations. In this study, hair samples were cut using various tools, including a scalpel, utility knife, and scissors. SEM images were then obtained, and the morphological differences between tool-cut hair and hair severed by tensile force were compared and analyzed. The results indicated that tool-cut hair exhibited distinct striation marks on the cross-section, with the cuticle layer shifting in the direction of tool insertion, leaving scratch-like marks. In contrast, hair that was forcibly pulled apart showed no tool marks and displayed an irregular fracture pattern. These findings contribute to the quantitative and qualitative assessment of hair cut marks in forensic investigations and provide fundamental data for expanding the application of morphological hair analysis in criminal cases.

현대국어의 ‘오히려’는 [도리어]와 [차라리]의 의미를 나타내는 부사이다. 15세기에도 ‘오히려’라는 형태가 존재했지만 그 의미는 현재와 큰 차이를 보인다. 15세기 ‘오히려’는 [도리어/반대로]와 [여전히/그대로]의 의미 외에, ‘하물며(況) 구문’과 높은 비율로 공기하면서 [심지어]의 의미로도 사용된 부사임을 새롭게 밝혔다. 동시에 현대국어 ‘오히려’가 [차라리]의 의미까지 표현했던 것과 달리, 15세기에서는 아직 이 의미까지는 나타내지 못한 것으로 조사되었다. 한편 15세기 ‘오히려’의 동의어로 알려진 ‘ᄉᆞᆫᄌᆡ’와의 비교를 통해서도 이 형태의 의미를 분명히 하고자 했다. ‘ᄉᆞᆫᄌᆡ’는 [도리어], [여전히]의 의미는 공유해도 [심지어]의 의미는 거의 나타내지 못했다. 또한 ‘오히려’는 [도리어]의 의미로 쓰인 예가 압도적으로 많았던 반면, ‘ᄉᆞᆫᄌᆡ’는 [여전히]의 의미로 더 자주 쓰이고 있어, 두 형태의 그 중심 의미와 의미 부담 비중이 서로 달랐던 것으로 파악되었다. 마지막으로 ‘오히려’는 ‘오힐-’에 ‘-여’가 결합한 형태로 보는 것보다 ‘오히-’의 영파생부사 ‘오히’에 부사파생접미사 ‘-려’로 분석하는 것이 더 타당함을 밝혔다.

In the 15th century, ‘오히려(Ohiryŏ)’ is an adverb that has the meaning of ‘even’, ‘rather’, and ‘still’. Unlike modern Korean ‘오히려(ohiryŏ)’, the meaning of ‘차라리(Chalali)’ was not expressed in the 15th century. And the synonym ‘ᄉᆞᆫᄌᆡ(Sanjae)’ could hardly express the meaning of ‘even’ even if the meanings of ‘rather’, and ‘still’ were shared. In addition, while ‘오히려(Ohiryŏ)’ is over- whelmingly used as the meaning of ‘rather’, ‘ᄉᆞᆫᄌᆡ(Sanjae)’ is often used as the meaning of ‘still’, so even if they are a synonym, the central meaning seem to have been different. Finally, ‘오히려(Ohiryŏ)’ can be analyzed in the form of a combination of the adverb ‘오히(ohi)’ and the adverbial suffix ‘-려(ryŏ)’.

In this study, we utilized the historical text morphological analysis tool ‘UTagger-Hunminjeongeum TCM’ to conduct a morphological corpus analysis of The Letters Excavated from the Tomb of the Suncheon Kim Clan. Based on this corpus, we examined the characteristics and trends of gendered language in male and female letters using quantitative methods. Chapter 2 explores indicators for the quantitative analysis of historical texts, dividing them into grammatical and lexical characteristics. Chapter 3 provides an overview of the structure of ‘UTagger-Hunminjeongeum TCM,’ the preprocessing steps for tagged data, and the data extraction and analysis process. Chapter 4 examines the patterns and characteristics of male and female letters based on the extracted data. Through this process, we identified gender-based linguistic preferences within linguistic indicators observed in the morphological analysis corpus of The Letters Excavated from the Tomb of the Suncheon Kim Clan.