질의문 키워드의 가중치 부여 방법을 이용한 정답 문서 순위화 시스템

강유환,안영민,서영훈 충북대학교 컴퓨터정보통신 연구소 2005 컴퓨터정보통신연구 Vol.13 No.1

본 논문에서는 자연어로 된 질의문에 대한 키워드 추출 및 가중치 부여 방법을 이용하여 정답 문서를 상위로 순위화하는 방법을 제안한다. 정답 문서를 상위로 순위화하기 위해 자연어 질의문과 정답 문서에서 나타나는 키워드의 특성을 분석하고, 색인어로써 중요한 역할을 할 수 있는 키워드를 다양한 형태로 추출한 후, 추출된 키워드에 차별화된 가중치를 부여한다. 끝으로 문서에 나타난 키워드의 출현 횟수와 문서 가중치를 이용하여 문서를 순위화한다. 실험에서는 20개의 질의문에 대해 상위 5위 안에 들어 있는 정답 문서의 수를 측정하였다. 실험결과 상위 5위 안에 들어 있는 정답 문서의 평균 개수는 3.05개로 키워드의 출현횟수와 키워드의 근접상을 이용한 방법보다 높은 성능 향상을 보였다. This paper describes a method of weighting keywords in the question that is written in a natural language for ranking answer documents. We extract keywords from the question using characteristic of question and keywords that are used in answer documents. And we assign differentiated weight to each keyword. Finally, we rank documents in order that answer documents can be ranked in high position using frequency of keywords and weight of documents. In experiment, we tested the number of answer documents that are ranked in top 5 for 20 questions. Experimental results show that the average number of answer documents that are ranked in top 5 was 3.05 and the accuracy of the proposed method was improved comparing with the existing method using frequency and proximity of keywords.

迎日灣의 ECOSYSTEM 硏究 2. 기후적 要因 : 氣溫, 日射量, 증발량, 바람, 강우량

朴相來,金基台,崔銀珠,李東澈,安由煥,尹伊鏞 嶺南大 海洋科學硏究所 1988 Marine Nature Vol.1 No.1

In order to understand phytoplankton ecology in the Gulf of Yeong-il, climatic factors such as air-temperature, solar-radiation, evaporation, direction and strength of winds, and precipitation are studied from January 1982 to December 1985. Annual average of air-temperature during the four-years period is about 13.6C, ranging from 1.7C (January) to 26.5C (August). The maximum value of solarradiation is measured in spring (520MJm²in May), the minimun : in winter (62 MJm²in December, January or February), and the total annual value : about 4987MJm². Annual average of evaporation is about 1200mm (Maximum: 169 in May : Minimun: 56 in February), close to precipitation (about 1050mm). Annual average of strength of "North Winds" (350°-90°)is about 550m·sec for 156 frequences during 191 days, "West Winds" (170°-340°) : 511m·sec¹for 185 frequences during 148 days, and "South Winds" (100°-160°) about 61m·sec¹for frequences during 24 days. "N" winds is about 30% more violant than "W" winds, but both winds are influential in terms of water movements in the Gulf.

迎日灣의 ECOSYSTEM 硏究 3. 水溫

尹伊鏞,朴相來,李東澈,金基台,崔銀珠,安由煥 嶺南大 海洋科學硏究所 1988 Marine Nature Vol.1 No.1

In order to observe hydrological characteristics in the Gulf of Yoeng-il, measurements of water temperature have been carried out from February to August 1983, and from August 1985 to December 1986. Surface water temperatures are closely linked to air temperature and solar-radiation for all littoral stations. Seasonal and annual variations of all horizontal stations are very similar to one another. Vertical variations of water temperatures for the stations located along the central-line of the Gulf are characteristic (and observed) as thermoclines and thermal stratification in the hottest summer-time (August). It is interesting to note that water column temperature of July and September are rather warmer than August, on account of a cold water mass coming from NE direction. For example, surface temperature in June 1982 was above 19℃, but the water below a depth of about 30m drastically drops to 4℃. It is indicative of invasion of water mass from cold current or deep sea.

Derivation and Comparison of Narrow and Broadband Algorithms for the Retrieval of Ocean Color Information from Multi-Spectral Camera on Kompsat-2 Satellite

Ahn, Yu-Hwan,Shanmugam, Palanisamy,Ryu, Joo-Hyung,Moon, Jeong-Eom The Korean Society of Remote Sensing 2005 大韓遠隔探査學會誌 Vol.21 No.3

The present study aims to derive and compare narrow and broad bandwidths of ocean color sensor’s algorithms for the study of monitoring highly dynamic coastal oceanic environmental parameters using high-resolution imagery acquired from Multi-spectral Camera (MSC) on KOMPSAT-2. These algorithms are derived based on a large data set of remote sensing reflectances ($R_{rs}$) generated by using numerical model that relates $b_b/(a + b_b)$ to $R_{rs}$ as functions of inherent optical properties, such as absorption and backscattering coefficients of six water components including water, phytoplankton (chl), dissolved organic matter (DOM), suspended sediment (SS) concentration, heterotropic organism (he) and an unknown component, possibly represented by bubbles or other particulates unrelated to the first five components. The modeled $R_{rs}$ spectra appear to be consistent with in-situ spectra collected from Korean waters. As Kompsat-2 MSC has similar spectral characteristics with Landsat-5 Thematic Mapper (TM), the model generated $R_{rs}$ values at 2 ㎚ interval are converted to the equivalent remote sensing reflectances at MSC and TM bands. The empirical relationships between the spectral ratios of modeled $R_{rs}$ and chlorophyll concentrations are established in order to derive algorithms for both TM and MSC. Similarly, algorithms are obtained by relating a single band reflectance (band 2) to the suspended sediment concentrations. These algorithms derived by taking into account the narrow and broad spectral bandwidths are compared and assessed. Findings suggest that there was less difference between the broad and narrow band relationships, and the determination coefficient $(r^2)$ for log-transformed data [ N = 500] was interestingly found to be $(r^2)$ = 0.90 for both TM and MSC. Similarly, the determination coefficient for log-transformed data [ N = 500] was 0.93 and 0.92 for TM and MSC respectively. The algorithms presented here are expected to make significant contribution to the enhanced understanding of coastal oceanic environmental parameters using Multi-spectral Camera.

ChlorophyII and suspended sediment specific absorption coefficient in the sea

Ahn, Yu Hwan,Moon, Jeong Eon 대한원격탐사학회 1998 International Symposium on Remote Sensing Vol.14 No.1

Absorption coefficient per mass unit of particles, specific absorption coefficient, is one of main parameters in developing algorithms for ocean color remote sensing. Specific absorption coefficient of chlorophyll (a*_(ph)) and suspended sediment (SS) were analyzed by $quot;wet filter technique$quot; and $quot;Kishino method$quot; for data sets observed in the Yellow and Mediterranean Seas. A new data-recovering method for the filter technique was also developed using spectrum slopes. This method recovered the baseline of spectrum that was often missed in the Kishino method. High a^*_(Ph)( λ ) values in the oligotrophic Mediterranean Sea and low values in the Yellow Sea were observed, spanning over the range of 0.02 to 0.12 ㎡/mg, at the chlorophyll maximum absorption wavelength 440nm. The empirical relationship between a^*_(Ph) and chlorophyll concentration was found to fit a power function, which was slightly different from that proposed by Bricaud et al. (1995). Absorption specific coefficients for suspended sediment (a^*_(ss)) didn't show any relationship with concentrations of suspended sediment. However, the average value of a^*_(ss) at 440nm was close to the specific absorption coefficient of soil (loess) measured by Ahn (1990). The more-pronounced variability of a^*_(ss) than a^*_(Ph) perhaps can explain more wide range of size-distribution for SS, which were determined by their specific gravity and agitation of water mass in the sea surface.

The Present Status of Ocean Color Remote Sensing in Korea

Ahn, Yu-Hwan 영남대학교 해양과학연구소 2000 국제심포지움 일정 및 발표논문집 Vol.2000 No.1

해색 원격탐사에서 가장 큰 문제가 되는 것은 다양한 광학적 특성을 갖는 해양에서 어떻게 하면 일차생산량의 지표인 클로로필 농도를 정확하게 얻을 수 있는가하는 것이다. 한국 해양연구소(KORDI)에서는 이러한 기술적 문제 해결과 한반도 주변의 해양의 생물학적 생태를 모니터링을 위한 도구로 사용하여 왔다. 지금까지는 주로 NOAA, CZCS, 그리고 SeaWiFS를 사용하여 기존의 생물광학 알고리즘을 개선하고, 현재는 CASE-II 해수 알고리즘, 탁도 등 해양환경분석 알고리즘을 개발하고 있다. 국립수산진흥원(NFRDI)에서는 이들 자료를 활용하여 현장의 어민들에게 매일매일 자료를 어장 정보를 위한 서비스와 위성 분석자료의 검증을 위한 현장 관측을 수행하고 있다. 현재 SeaWiFS 수신시스템을 갖추고 있는 곳은 한국해양연구소, 국립수산진흥원, 그리고 기타 1곳에 갖추고 있다. 그러나 대학에서는 한곳도 없는 것이 특징이다. 이 사실은 한국 내 Ocean color 분양의 자료 이용자나 알고리즘 개발분야에 인력이 그렇게 많지 않음을 시사한다. 그럼에도 불구하고 1999년 12월 국내에서는 육상 관측센서인 ECO를 주 탑재체로 하고, 해색센서(OSMI)가 탑재된 아리랑 위성 1호를 개발하여 발사하였다. OSMI의 bands 특성은 SeaWiFS와 거의 유사하나 band 수가 6개 뿐이다. 현재는 이들 자료를 활용을 위한 CAL/VAL 사업이 KORDI/NFRDI에서 동시에 수행 중에 있으며 자료의 해양분야 활동은 거의 없는 실정이다. 미래의 해색 기술연구를 위하여 현재 KORDI의 해양과학 및 해색원격탐사 개발팀은 남해에서 연중 발생하는 적조의 조기예보를 위한 정밀 적조 알고리즘을 개발하고 있으며, 앞으로 몇몇 적조의 종까지 구분이 가능한 기초연구를 수행하고 있다. The initial difficulty in ocean color remote sensing lies in the concentration determination of various ocean constituents (phytoplankton, total seston, etc.) found in surface ocean wafers. The Korea Ocean Research and Development Institute (KORDI) carries out most of researches on ocean color satellite technique. KORDI has utilized and improved earlier techniques for analyzing sea surface temperature (SST), chlorophyll concentrations as well as water turbidity using CZCS, NOAA and SeaWiFS data. The National Fisheries Research and Development Institute (NFRDI) of Korea provides information on daily distribution of SST to fishermen of the area of concerns. Receiving stations for SeaWiFS are limited to only KORDI and NFRDI in Korea. The first Korean Multi-purpose Satellite (KOMSAT) is in operation with two sensors, ocean color (OSMI) and land (EOC) sensors. OSMI band characteristics of KOMSAT are very similar to those of the SeaWiFS of NASA. The ocean color-group at KORDI has been developing algorithms for the accurate detection of initialization of red-tides in order to predict in advance. At the present, the group is working on an optical calibration and validation of OSMI sensor

New Methods for Correcting the Atmospheric Effects in Landsat Imagery over Turbid (Case-2) Waters

Ahn Yu-Hwan,Shanmugam P. The Korean Society of Remote Sensing 2004 大韓遠隔探査學會誌 Vol.20 No.5

Atmospheric correction of Landsat Visible and Near Infrared imagery (VIS/NIR) over aquatic environment is more demanding than over land because the signal from the water column is small and it carries immense information about biogeochemical variables in the ocean. This paper introduces two methods, a modified dark-pixel substraction technique (path--extraction) and our spectral shape matching method (SSMM), for the correction of the atmospheric effects in the Landsat VIS/NIR imagery in relation to the retrieval of meaningful information about the ocean color, especially from Case-2 waters (Morel and Prieur, 1977) around Korean peninsula. The results of these methods are compared with the classical atmospheric correction approaches based on the 6S radiative transfer model and standard SeaWiFS atmospheric algorithm. The atmospheric correction scheme using 6S radiative transfer code assumes a standard atmosphere with constant aerosol loading and a uniform, Lambertian surface, while the path-extraction assumes that the total radiance (L/sub TOA/) of a pixel of the black ocean (referred by Antoine and Morel, 1999) in a given image is considered as the path signal, which remains constant over, at least, the sub scene of Landsat VIS/NIR imagery. The assumption of SSMM is nearly similar, but it extracts the path signal from the L/sub TOA/ by matching-up the in-situ data of water-leaving radiance, for typical clear and turbid waters, and extrapolate it to be the spatially homogeneous contribution of the scattered signal after complex interaction of light with atmospheric aerosols and Raleigh particles, and direct reflection of light on the sea surface. The overall shape and magnitude of radiance or reflectance spectra of the atmospherically corrected Landsat VIS/NIR imagery by SSMM appears to have good agreement with the in-situ spectra collected for clear and turbid waters, while path-extraction over turbid waters though often reproduces in-situ spectra, but yields significant errors for clear waters due to the invalid assumption of zero water-leaving radiance for the black ocean pixels. Because of the standard atmosphere with constant aerosols and models adopted in 6S radiative transfer code, a large error is possible between the retrieved and in-situ spectra. The efficiency of spectral shape matching has also been explored, using SeaWiFS imagery for turbid waters and compared with that of the standard SeaWiFS atmospheric correction algorithm, which falls in highly turbid waters, due to the assumption that values of water-leaving radiance in the two NIR bands are negligible to enable retrieval of aerosol reflectance in the correction of ocean color imagery. Validation suggests that accurate the retrieval of water-leaving radiance is not feasible with the invalid assumption of the classical algorithms, but is feasible with SSMM.

High Resolution Visible and IR sensor on the Geostationary Satellite

Ahn, Yu Hwan 대한원격탐사학회 2001 International Symposium on Remote Sensing Vol.17 No.1

The High Resolution Visible and IR Sensor (HiRVIS; an assumed name) is a future ocean color and IR sensor to be loaded on the Korean Multifunctional GEO Satellite (in 2008) for the purpose of obtaining multispectral visible and infra-red images of the ocean (land) surface at high repeat observation. These images will be used to monitor and study processes which can vary rapidly in time in the ocean surface. HiRVIS has a spatial resolution of 250m x 250m (for visible bands) and an image size of 500km x 500km. It has the capability of providing an image of eleven spectral bands (visible NIR 8, IR 3) every 30minute, which greatly expands the capability of OSMI or SeaWiFS in spatial and temporal resolution. HiRVIS is designed to remedy the coverage and revisiting rate contractions imposed by polar orbiting platform. The high frequency observing rate can also increase of cloud-free area through cloud filtering and composition of multiple images. The images are needed in management of coastal zone resources; and mitigation of marine hazards such as inflow of turbid fresh water in the ocean, up-welled cold water, and harmful algal bloom etc.. Data will be processed in near real time by KORDI and NFRDI. Research data products will be available through KORDI.

New Methods for Correcting the Atmospheric Effects in Landsat Imagery over Turbid (Case-2) Waters

Yu Hwan Ahn,P. Shanmugam 大韓遠隔探査學會 2004 大韓遠隔探査學會誌 Vol.20 No.5

Atmospheric correction of Landsat Visible and Near Infrared imagery (VIS/NIR) over aquatic environment is more demanding than over land because the signal from the water column is small and it carries immense information about biogeochemical variables in the ocean. This paper introduces two methods, a modified dark-pixel substraction technique (path-extraction) and our spectral shape matching method (SSMM), for the correction of the atmospheric effects in the Landsat VIS/NIR imagery in relation to the retrieval of meaningful information about the ocean color, especially from Case-2 waters (Morel and Prieur, 1977) around Korean peninsula. The results of these methods are compared with the classical atmospheric correction approaches based on the 6S radiative transfer model and standard SeaWiFS atmospheric algorithm. The atmospheric correction scheme using 6S radiative transfer code assumes a standard atmosphere with constant aerosol loading and a uniform, Lambertian surface, while the path-extraction assumes that the total radiance (L(TOA)) of a pixel of the black ocean (referred by Antoine and Morel, 1999) in a given image is considered as the path signal, which remains constant over, at least, the sub scene of Landsat VIS/NIR imagery. The assumption of SSMM is nearly similar, but it extracts the path signal from the L(TOA) by matching-up the in-situ data of water-leaving radiance, for typical clear and turbid waters, and extrapolate it to be the spatially homogeneous contribution of the scattered signal after complex interaction of light with atmospheric aerosols and Raleigh particles, and direct reflection of light on the sea surface. The overall shape and magnitude of radiance or reflectance spectra of the atmospherically corrected Landsat VIS/NIR imagery by SSMM appears to have good agreement with the in-situ spectra collected for clear and turbid waters, while path-extraction over turbid waters though often reproduces in-situ spectra, but yields significant errors for clear waters due to the invalid assumption of zero water-leaving radiance for the black ocean pixels. Because of the standard atmosphere with constant aerosols and models adopted in 6S radiative transfer code, a large error is possible between the retrieved and in-situ spectra. The efficiency of spectral shape matching has also been explored, using SeaWiFS imagery for turbid waters and compared with that of the standard SeaWiFS atmospheric correction algorithm, which fails in highly turbid waters, due to the assumption that values of water-leaving radiance in the two NIR bands are negligible to enable retrieval of aerosol reflectance in the correction of ocean color imagery. Validation suggests that accurate the retrieval of water-leaving radiance is not feasible with the invalid assumption of the classical algorithms, but is feasible with SSMM.

Development of Suspended Particulate Matter Algorithms for Ocean Color Remote Sensing

Yu Hwan Ahn,Jeong Eun Moon,Sonia Gallegos 大韓遠隔探査學會 2001 大韓遠隔探査學會誌 Vol.17 No.4

We developed a CA5E-Ⅱ water model that will enable the simulation of remote sensing reflectance(Rrs) at the coastal waters for the retrieval of suspended sediments (SS) concentrations from satellite imagery. The model has six components which are: water, chlorophyll, dissolved organic matter (DOM), non-chlorophyllous particles (NC), heterotrophic microorganisms and an unknown component, possibly represented by bubbles or other particulates unrelated to the five first components. We measured Rrs concentration of SS and chlorophyll, and absorption of DOM during our field campaigns in Korea. In addition, we generated Rrs from different concentrations of SS and chlorophyll, and various absorptions of DOM by random number functions to create a large database to test the model. We assimilated both the computer generated parameters as well as the in-situ measurements in order to reconstruct the reflectance spectra. We validated the model by comparing model-reconstructed spectra with observed spectra. The estimated Rrs spectra were used to (1) evaluate the performance of four wavelengths and wavelengths ratios for accurate retrieval of SS. 2) identify the optimum band for SS retrieval, and 3) assess the influence of the SS on the chlorophyll algorithm. The results indicate that single bands at longer wavelengths in visible yield better results than commonly used channel ratios. The wavelength of 625nm is suggested as a new and optimal wavelength for SS retrieval. Because this wavelength is not available from SeaWiFS, 555nm is offered as an alternative. The presence of SS in coastal areas can lead to overestimation chlorophyll concentrations greater than 20-500%.