• The Journal of Korean Association of Computer Education
• /
• v.10 no.3
• /
• pp.67-73
• /
• 2007

In this paper, we review essential constituents of a dialogue interface system and propose practical methods to implement the each constituent. The implemented system consists of a discourse manager, an intention analyzer, a named entity recognizer, a SQL query generator, and a response generator. In the progress of implementation, the intention analyzer uses a maximum entropy model based on statistics because the domain dependency of the intention analyzer is comparatively low. The others use a simple pattern matching method because they needs high domain portability. In the experiments in a schedule arrangement domain, the implemented system showed the precision of 88.1% in intention analysis and the success rate of 83,4% in SQL query generation.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.22 no.1
• /
• pp.26-46
• /
• 2020

An irrigated-maize agroecosystem is viewed as an open thermodynamic system upon which solar radiation impresses a large gradient that moves the system away from equilibrium. Following the imperative of the second law of thermodynamics, such agroecosystem resists and reduces the externally applied gradient by using all means of this nature-human coupled system acting together as a nonequilibrium dissipative process. The ultimate purpose of our study is to test this hypothesis by examining the energetics of agroecosystem growth and development. As a first step toward this test, we employed the eddy covariance flux data from 2003 to 2014 at the AmeriFlux NE1 irrigated-maize site at Mead, Nebraska, USA, and analyzed the energetics of this agroecosystem by scrutinizing its radiation, energy and entropy exchange. Our results showed: (1) more energy capture during growing season than non-growing season, and increasing energy capture through growing season until senescence; (2) more energy flow activity within and through the system, providing greater potential for degradation; (3) higher efficiency in terms of carbon uptake and water use through growing season until senescence; and (4) the resulting energy degradation occurred at the expense of increasing net entropy accumulation within the system as well as net entropy transfer out to the surrounding environment. Under the drought conditions in 2012, the increased entropy production within the system was accompanied by the enhanced entropy transfer out of the system, resulting in insignificant net entropy change. Drought mitigation with more frequent irrigation shifted the main route of entropy transfer from sensible to latent heat fluxes, yielding the production and carbon uptake exceeding the 12-year mean values at the cost of less efficient use of water and light.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2001.06a
• /
• pp.177-180
• /
• 2001

본 논문에서는 신경망을 이용한 내용 기반 이미지 분류 방법을 제안한다. 분류 대상이미지는 인터넷상의 다양한 이미지들 중 오브젝트 이미지이대 웹 에이전트를 통하여 획득하고 정규화 과정을 거친다. 획득한 이미지를 분류하기 위한 특징은 웨이블릿 변란 후 추출된 질감 특징이다. 추출된 질감 특징을 이용하여 학습패턴을 생성하고 신경망을 학습한다. 그리고 구성된 신경망 분류기로 이미지를 분류한다. 본 연구에서는 다양한 질감 특징들 중에서 대비(contrast), 에너지(energy), 엔트로피(entropy)를 이용하여 특징을 추출한다. 실험에 사용한 데이터는 30종류에 대하여 각각 10개씩, 300개의 이미지들을 학습 데이터, 테스트 데이터로 사용하여 구성된 분류기의 인식률을 실험하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.3
• /
• pp.423-431
• /
• 1997

A new decision tree generator MEC is proposed in this paper, which uses the difference of multi-base entropy as a consistent criterion for discretization and selection of attributes. To evaluate the performance of the proposed generator, it is compared to other generators which use criteria based on entropy and adopt different discretization styles. As an experimental result, it is shown that the proposed generator produces the most efficient classifiers, which have the least number of leaves at the same error rate, regardless of whether attribute values constituting the training set are discrete or continuous.

• The Journal of the Korea Contents Association
• /
• v.16 no.8
• /
• pp.160-168
• /
• 2016

The classification and evaluation of cooking that is being published on the internet are presented without scientific criteria based on individual subjective factors. In this paper, we objectified the degree of cooking difficulty based on the information entropy. And we measured the similarity by calculating the common entropy between recipes and constructed a social network based on the recipe similarity. As a result of measuring the cooking difficulty, 'Dongtae Haemul-jjim' (Korean) and 'Vegetarian Lasagna' (Italy) are the most difficult recipes and 'Gochu-jang' (Korean) and 'Tofu steak' (Italy) are the easiest recipes. Through the recipe network, the similarity between Korean and Asian cooking is higher than Western cuisine. We showed a similar recipe to a particular cooking, the group of similar recipes, and reasonable schedule when preparing the menu from the viewpoint of ease of cooking.

• The KIPS Transactions:PartB
• /
• v.18B no.2
• /
• pp.67-72
• /
• 2011

Single nucleotide polymorphisms (SNPs) are a very important tool for the study of human genome structure. Cluster analysis of the large amount of gene expression data is useful for identifying biologically relevant groups of genes and for generating networks of gene-gene interactions. In this paper we compared the clusters of SNPs within asthma group and normal control group obtained by using hierarchical cluster analysis method with entropy distance. It appears that the 5-cluster collections of the two groups are significantly different. We searched the best set of SNPs that are useful for diagnosing the two types of asthma using representative SNPs of the clusters of the asthma group. Here support vector machines are used to evaluate the prediction accuracy of the selected combinations. The best combination model turns out to be the five-locus SNPs including one on the gene ALOX12 and their accuracy in predicting aspirin tolerant asthma disease risk among asthmatic patients is 66.41%.

• Journal of Korean Society for Geospatial Information Science
• /
• v.14 no.4 s.38
• /
• pp.19-26
• /
• 2006

This study assessed the accuracies of the ground information extracting methods from the LiDAR data. Especially, it compared two kinds of method, one of them is using directly the raw LiDAR data which is point type vector data and the other is using changed data to DSM type as the normal grid type. The methods using Local Maxima and Entropy methods are applied as a former case, and for the other case, this study applies the method using edge detection with filtering and the generated reference surface by the mean filtering. Then, the accuracy assessment are performed with these results, DEM constructed manually and the error permitted limit in scale of digital map. As a results, each DEM mean errors of methods using edge detection with filtering, reference surface, Local Maxima and Entropy are 0.27m, 2.43m, 0.13m and 0.10m respectively. Hence, the method using entropy presented the highest accuracy. And an accuracy from a method directly using the raw LiDAR data has higher accuracy than the method using changed data to DSM type relatively.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.1
• /
• pp.59-65
• /
• 1999

In this paper, we design a entorpy encoder of HDTV/DTV encoder blocks on the basis of MPEG-II. The designed entropy encoder outputs its bit stream at 9Mbps bit rate inserting zero-stepping block to protect the depletion of buffer in case that the generated bit stream is stored in buffer and uses not only PROM bit combinational circuit to solve the problem of critical path, and packer block, one of submerge, is designed to packing into 24 bit unit using barrel shifter, and it is constructed to blocks of header information encoder, input information delay, submerge, and buffer control. Designed circuits is verified by VHDL function simulation, as a result of performing P&R with Gate compiler that apply $0.8{\mu}m$ Gate Array specification, pin and gate number of total circuits has been tested to each 235 and about 120,000.

• Journal of Korea Multimedia Society
• /
• v.3 no.2
• /
• pp.201-208
• /
• 2000

In this paper We describe an algorithm which is devised for 4he partition o# the input space and the generation of fuzzy rules by the fuzzy entropy and tested with the time series prediction problem using Mackey-Glass chaotic time series. This method divides the input space into several fuzzy regions and assigns a degree of each of the generated rules for the partitioned subspaces from the given data using the Shannon function and fuzzy entropy function generating the optimal knowledge base without the irrelevant rules. In this scheme the basic idea of the fuzzy neural network is to realize the fuzzy rules base and the process of reasoning by neural network and to make the corresponding parameters of the fuzzy control rules be adapted by the steepest descent algorithm. The Proposed algorithm has been naturally derived by means of the synergistic combination of the approximative approach and the descriptive approach. Each output of the rule's consequences has expressed with its connection weights in order to minimize the system parameters and reduce its complexities.

• The Korean Journal of Applied Statistics
• /
• v.17 no.1
• /
• pp.135-144
• /
• 2004

We propose a reproducibility (validity) assessment procedure of K-means cluster analysis by randomly partitioning the data set into three parts, of which two subsets are used for developing clustering rules and one subset for testing consistency of clustering rules. Also, as an alternative to Rand index and corrected Rand index, we propose an entropy-based consistency measure between two clustering rules, and apply it to determination of the number of clusters in K-means clustering.