한국어 연속음성에서의 조사 및 어미 인식에 관한 연구

송민석,이기영 한국음성과학회 1999 음성과학 Vol.6 No.1

This study proposes a method of recognizing postpositions and suffixes in Korean spoken language, using prosodic information. We detect grammatical boundaries automatically at first, by using prosodic information of the accentual phrase, and then we recognize grammatical function words by backward-tracking from the boundaries. The experiment employs 300 sentential speech data of 10 men's and 5 women's voice spoken in standard Korean, in which 1080 accentual phrases and II postpositions and suffixes are included. The result shows the recognition rate of postpositions in two cases. In one case in which only correctly detected boundaries are included, the recognition rate is 97.5%, and in the other case in which all detected boundaries are included, the recognition rate is 74.8%. Keywords: accentual phrase, backward-tracking, continuous speech

Poly(urea-urethane) 자기복원 코팅층을 가진 도장 보호필름 물성 연구

송민석 한국접착및계면학회 2023 접착 및 계면 Vol.24 No.2

Recently, the application of paint protection films (PPFs) for automobiles having a self-recovery coating layer has been grown up. In this study, we report the evaluation results on the basic physical properties of a poly(vinyl chloride)- based PPF containing poly(urea-urethane) hybrid self-recovery coating layer. Depending on the main chemical composition and the thickness of poly(urea-urethane)-based coating layer for PPF, the self-recovery performance by an optical microscope and the stain resistance through color difference value are measured. To improve the surface properties and show its easy-cleaning effect against the polluted things, silicone-modified polyacrylate is introduced to the self-recovery coating composition. The contact angle of water on the coated surface is confirmed to show its hydrophobic surface. Finally, accelerated weathering test of paint protection film with poly(urea-urethane) hybrid coating layer is performed to check the possibility of discoloration and deformation due to long-term exposure on harsh condition.

高等學校 理念敎育의 實態와 改善 方案

宋珉錫 한국사회과교육연구회 1990 사회과교육 Vol.- No.23

텔레메틱스 환경에서 다자간 회의 시스템을 위한 음량균일화알고리즘

송민석,김형국,류상현 한국ITS학회 2013 한국ITS학회 학술대회 Vol.2013 No.1

대용량 저장 서버에서 디스크 소모 전력 감소 기법에 관한 연구 동향

송민석,김만종,양아론 한국정보처리학회 2009 정보처리학회지 Vol.16 No.6

고압 유체 시료의 pH 및 알칼리도 측정 방법 : 가상 시료를 활용한 실용성 평가

송민석,문수현,채기탁,방준환,Minseok Song,Soohyeon, Moon,Gitak Chae,Jun-Hwan Bang 한국지하수토양환경학회 2024 지하수토양환경 Vol.29 No.1

As part of monitoring technology aimed at verifying the stability of CO₂ geologic storage and mitigating concerns about leakage, a method for measuring the pH and alkalinity of high-pressure fluid samples was established to obtain practical technology. pH measurement for high-pressure samples utilized a high-pressure pH electrode, and alkalinity was measured using the Gran titration method for samples collected with a piston cylinder sampler (PCS). Experimental samples, referencing CO₂-rich water and CO₂ geologic storage studies, were prepared in the laboratory. The PCS controls the piston, preventing CO₂ degassing and maintaining fluid pressure, allowing mixing with KOH to fix dissolved CO₂. Results showed a 6.1% average error in high-pressure pH measurement. PCS use for sample collection maintained pressure, preventing CO₂ degassing. However, PCS-collected sample alkalinity measurements had larger errors than non-PCS measurements, limiting PCS practicality in monitoring field settings. Nevertheless, PCS could find utility in preprocessing for carbon isotope analysis and other applications. This research not only contributes to the field of CCS monitoring but also suggests potential applications in studies related to natural analogs of CCS, CO₂-rock interaction experiments, core flooding experiments, and beyond.

음운론을 이용한 영어 발음 교육 : Based on Phonological Knowledge:The Case of Liquids/r/and/l/ 국어/ㄹ/과 영어/r,l/의 예

宋旼錫 관동대학교 1996 關大論文集 Vol.24 No.2

Teaching English pronunciation in Korea puts emphasis only on phonetic characteristics of each sound. For example, in the case of liquids, Korean has the lateral [l] and English has the lateral [l] and the retroflex [r]. One the ground that Karen has no retroflex [r], only the pronunciation of [r] is brought into focus and the pronounced of [l] is neglected in teaching English pronunciation. But Korean /l/ is pronounced as ether [r] or [l] according to the position where it occurs. The problem lies on the fact that the position where the Korean [l] is pronounced is different form that of the English [l]. The lateral [l] cannot be pronounced in the syllable initial position in Korean, while it can be freely pronounced in that position in English. Therefore, it is important to teach not only the pronunciation of [r] but the pronunciation [l] in the syllable initial position. Thus, in teaching the pronunciation of foreign languages, it is necessary that the phonological structure of the target language should be compared with that of the native language and the errors originating from the difference of the phonological structures should be corrected. The purpose of this paper is to bring into focus not only the phonetic characteristic of each sound but also the importance of the phonological structures, in teaching the pronunciation of foreign languages. For this purpose, this paper compares the phonological structure of English with that of Korean and examines what must be emphasized teaching English pronunciation, taking the liquids as an example.

프로세스 마이닝을 활용한 생산 공정 데이터 분석

송민석,정일교,조용주,조현제 한국경영과학회 2012 한국경영과학회 학술대회논문집 Vol.2012 No.5

Exploratory evaluation of the role of cardiac troponin on the clinical outcome of patients visiting emergency department with or without chronic kidney disease

송민석,강은진,김태림,박종은,이건탁,윤희,황성연,차원철,신태건,심민섭,조익준,최진호 대한응급의학회 2022 대한응급의학회지 Vol.32 No.6

Objective: Elevated levels of cardiac troponin in chronic kidney disease (CKD) patients admitted to the emergency department (ED) is not well understood and is often ignored. This study aimed to investigate the impact of cardiac troponin I (TnI) levels on the clinical outcome of patients visiting the ED with or without CKD. Methods: In this retrospective single-center cohort study, we enrolled patients visiting the ED without a diagnosis of coronary artery disease (CAD). Elevated cardiac TnI was defined as being ≥99th percentile of the normal population (Siemens ADVIA Centaur TnI-Ultra≥0.040 ng/mL). The clinical outcomes of patients with CKD stage≤2 and CKD stage ≥3 were compared. The primary endpoint was the 180-day all-cause death, including cardiovascular and non-cardiovascular deaths. Results: Among a total of 30,472 patients (median age, 61 years; male sex, 54.3%), elevated TnI was found in 4,377 patients (14.4%). There were 3,634 deaths (11.9%) including 584 cardiovascular (1.9%) and 3,050 non-cardiovascular deaths (10.0%). The risk of all-cause death increased in patients with elevated TnI in both CKD stage≤2 (hazard ratio [HR], 2.1; 95% confidence interval [CI], 1.9-2.3) and CKD stage≥3 (HR, 1.5; 95% CI, 1.4-1.7), and so did the risks of cardiovascular and non-cardiovascular death (HR, 1.2-4.7) (P<0.05, all). The association of elevated TnI with death risk was consistent in multivariate analyses and in most clinical subgroup analyses. Conclusion: Elevated TnI was associated with higher 180-day mortality irrespective of renal function among patients visiting the ED without documented CAD. CKD patients visiting the ED with elevated TnI may warrant additional evaluation or careful follow-up even without the presence of CAD.

구조보존규칙으로서의 영어 비음동화:[η]에 대한 해결책

宋旼錫 관동대학교 1995 關大論文集 Vol.23 No.2

The velar nasal [ŋ] is an allophone of /n/ in English. This fact is evidenced by distributional restrictions. That is, [ŋ] always appears before /g/ or /k/, while /m/ and /n/ occur freely. But in the theory of Lexical Phonology, the English Nasal Assimilation which derives the velar nasal [ŋ] must be a phonological rule which applies in the lexicon. For this rule must precede the morphological process suffixing '-er', as shown in the examples of "sing" and "singer" ("singer" is pronounced as [siŋ( )r], not as [siŋg( )r]). On the other hand, a lexical rule must be structure-preserving in the sense that if a feature is not distinctive in any context it may not play any role in the lexical phonology(Kiparsky, 1984). With regard to this principle of Structure Preservation, the above nasal assimilation has a problem because this process is not structure-preserving but must be lexical. To solve this problem, Borowsky(1986) says that the above nasal assimilation is applied at level 2 and that the principle of Structure Preservation turns off at level 2. This is an ad hoc solution. If this solution is accepted, the word of "concord" must be derived at level 2 while the words of "combine" and "concern" are derived at level 1 (independently motivated). This paper tries to solve this problem in the standpoint of the feature geometry theory. In the feature geometry, the process of spreading to an underspecified target is interpreted as structure-preserving. And the above nasal assimilation is a process of spreading to an underspecified target because the place feature of the phoneme /n/ is underspecified in the underlying representation. In the result, the English nasal assimilation is a structure-preserving process and can apply freely in the lexicon. As another desirable result, we can say that all lexical phonology is structure-preserving.