• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.2
• /
• pp.77-82
• /
• 2023

The Ministry of Agriculture, Food and Rural Affairs established the FarmMap, an digital map of agricultural land. In this study, using deep learning, we suggest the application of farm map reading to farmland such as paddy fields, fields, ginseng, fruit trees, facilities, and uncultivated land. The farm map is used as spatial information for planting status and drone operation by digitizing agricultural land in the real world using aerial and satellite images. A reading manual has been prepared and updated every year by demarcating the boundaries of agricultural land and reading the attributes. Human reading of agricultural land differs depending on reading ability and experience, and reading errors are difficult to verify in reality because of budget limitations. The farmmap has location information and class information of the corresponding object in the image of 5 types of farmland properties, so the suitable AI technique was tested with ResNet50, an instance segmentation model. The results of attribute reading of agricultural land using deep learning and attribute reading by humans were compared. If technology is developed by focusing on attribute reading that shows different results in the future, it is expected that it will play a big role in reducing attribute errors and improving the accuracy of digital map of agricultural land.

• Journal of The Korean Association of Information Education
• /
• v.26 no.5
• /
• pp.341-352
• /
• 2022

The wave of a knowledge and information society led by AI, Big Data, and so on is having an all-round impact on our way of life. Therefore the Ministry of Education is in a hurry to strengthen Digital Literacy, including AI and SW Education, by improving the curriculum that can cultivate basic knowledge and capabilities to respond to changes in the future society. It can be seen that establishing a foundation for cultivating Digital Literacy through all subjects and improving basic and in-depth learning in new technology fields such as AI linked to the information curriculum is an essential part for future society. However, research on each content for cultivating Digital and AI literacy is relatively active, while research on teaching and learning strategies is insufficient. Therefore in this study, a CT-based Digital and AI teaching and learning strategy that can foster that was developed and Delphi expert verification was conducted, and the final teaching and learning strategy was completed after evaluating instructor usability and analyzing learner effectiveness.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.239-248
• /
• 2023

The purpose of this study is to analyze the factors for service contracts that affect the successful bid price rate, focusing on the case of the country market. In the study, ordering organizations and bidders differentiated themselves from existing studies by analyzing service contracts that affect the successful bid price rate in a wide range of country markets. Comparative analysis of the awarding price ratio for services, this work provides a comparable result to the existing results in the previous literature. The analytical model used five independent variables such as budget, contract method, the days of the public notice, the awarding method, and the lowest awarding ratio. In the survey and analysis, big data was collected using text mining for service bids for Nara Market over the past 18 years and data was analyzed in a multi-dimensional way. The results of the analysis are as follows, (1) if budget does not determine the awarding price ratio. This is not the case in small amounts. (2) The contract method affects the awarding price ratio. (3) The days of the public notice increase, the awarding price ratio decrease. (4) the awarding method affects the awarding price ratio. (5) The lowest awarding ratio determines the awarding price ratio. Based on the results of empirical analysis, policy implications were sought.

• Journal of dental hygiene science
• /
• v.4 no.3
• /
• pp.91-95
• /
• 2004

This study aims to analyze realities of oral health and related factors, and establish Structural Equation Modeling. The subjects of study were 9,340 elderly over age 65 who took the health examination(the first) for the local insured which National Health Insurance Corporation carried out in the survey area mentioned below from January 2002 to December 2002. The areas surveyed were 4 big cities including Seongbuk-ku, Seoul, 5 medium cities including Wonjusi, Gangwon-do, and 5 Counties including Yeong deok County, Kyeongbuk. Considering location and the scale of population, firstly, big unit areas(metropolitan city, province) were selected according to convenience, secondly, low unit areas(city, county, district) were selected randomly. The subjects were the elderly who took all tests including an oral examination and filled in the questionnaire. Major results from analysis are as follows: 1. Review of Composition Conception Validity As a result of analyzing composition conception validity of SEM including posture test, urine test, blood test, habits of eating, drinking and smoking, oral symptoms, and oral health status, using fit index such as GFI, CFI, TLI, and RMSEA, all were within fit range and composition conception validity was recognized. 2. As a result of analyzing SEM to find the relationship between each factor and oral health status, it was confirmed that all factors except urine test affected oral health status and the synthetic SEM to explain it could be established. In result, we could verify that the elderly of rural areas who had lesser experience of visiting a clinic and oral prophylaxis had a higher rate of caries, missing teeth, and denture need, and drinking and smoking negatively affected the rate of caries, periodontal, and missing teeth. Also, periodontal diseases were observed from 43.2% of the total elderly and much from the lower age. Most of oral disease can be prevented by right oral health behavior. Therefore through oral health professionals from each district public health center of the nation, oral health education for the elderly about right eating habits and oral health care should be carried out systematically and policy change to increase access to dental service is required lest that visiting a dental clinic should be impossible or oral health behaviors such as oral prophylaxis and denture wearing should be neglected by economic, geographical barriers. Also, to establish SEM to explain the relationship between oral health status and systemic health, more accurate test methods and effective index development should be preceded. Because items developed by National Health Insurance Corporation applied to this study without alteration, structuring a model had the uppermost limit. Continual study seems to be needed.

• Journal of Distribution Science
• /
• v.9 no.1
• /
• pp.29-38
• /
• 2011

Since a distribution market was opened in 1993 after the conclusion of the Uruguay Round (UR), various new businesses emerged in the Korean retail industry, such as convenience shops, supermarkets, specialty stores, non-store marketing methods, and cyber shopping malls, in addition to traditional markets and department stores. Competition among these retail businesses has intensified. According to the National Statistical Office (NSO), the growth rate of the total retail industry has shown a 32% increase in the past 5 years. While department stores, supermarkets, specialty stores, and other non-store retailing venues have shown signs of stagnation at a growth rate of 20-30%, which is similar to the growth rate of the entire retail industry, convenience shops and non-store marketing have shown 60-70% growth over the same period. By comparison, the growth rate of cyber shopping malls has nearly tripled. When applying development aspects of retail businesses through the competition to the retail life cycle, mom-and-pop stores and traditional markets have already entered the decline phase as specialty stores reach their maturity phase and demonstrate their limit of growth. Department stores are now in the latter part of the growth phase, which is still considered to have some growth potential. Big super markets are still in the early part of their growth stage although they were introduced 20 years ago. Meanwhile, retail businesses such as convenience stores, supermarkets, mail order houses, and warehouse stores are entering the middle growth phase and are expected to continue with their quantitative growth. At a time when most retail businesses in Korea are in a state of development or in the full growth stage, what kind of new innovative retailing will appear and develop? Moreover, what growth engine will drive it? This study analyzes the appearance of innovative retailing and its development process by establishing a discussable consumer's choice model through the interlocking mutual behavior of differentiated competitiveness and consumers' choice based on an awareness of this issue. The analysis of the results of this study can be summarized as follows. First, if a new vacuum zone emerges at the retail market level, innovative retailing such as low price/low service or high price/high service will emerge simultaneously. Second, if the number of new businesses in the vacuum zone increases, this will create competitiveness among the businesses, and each retailer will develop raising of level. Third, if a new business that raises the level develops, competition between the new and the existing retailer will occur, and an assimilation process between the existing retailing and the new retailing will be unfolded. Fourth, each retailing will promote distribution innovations in order to break the frontier of the existing distribution technology, and other retailing will follow the innovator. On the basis of an analysis of the abovementioned results, this study presents the following three suggestions. First, responding to the consumer's decision-making process on the attributes of retail shops that promote differentiation in strategies, this study established a consumer's choice model that can be discussed in relation to changes in market share. Second, this study provided an analysis of the emerging and developmental processes of innovative retail businesses using a more precise logical structure on the basis of the consumer's choice model described in this study. Third, the development process of retail businesses discussed in this study presented retailing solutions regarding management aspects on how to compose a strong retail mix that can help retail businesses gain competitive advantages in the market.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.23 no.6
• /
• pp.469-484
• /
• 2021

Rock mass classification results have a great influence on construction schedule and budget as well as tunnel stability in tunnel design. A total of 3,526 tunnels have been constructed in Korea and the associated techniques in tunnel design and construction have been continuously developed, however, not many studies have been performed on how to assess rock mass quality and grade more accurately. Thus, numerous cases show big differences in the results according to inspectors' experience and judgement. Hence, this study aims to suggest a more reliable rock mass classification (RMR) model using machine learning algorithms, which is surging in availability, through the analyses based on various rock and rock mass information collected from boring investigations. For this, 11 learning parameters (depth, rock type, RQD, electrical resistivity, UCS, V_p, V_s, Young's modulus, unit weight, Poisson's ratio, RMR) from 13 local tunnel cases were selected, 337 learning data sets as well as 60 test data sets were prepared, and 6 machine learning algorithms (DT, SVM, ANN, PCA & ANN, RF, XGBoost) were tested for various hyperparameters for each algorithm. The results show that the mean absolute errors in RMR value from five algorithms except Decision Tree were less than 8 and a Support Vector Machine model is the best model. The applicability of the model, established through this study, was confirmed and this prediction model can be applied for more reliable rock mass classification when additional various data is continuously cumulated.

• Journal of Distribution Research
• /
• v.17 no.3
• /
• pp.25-57
• /
• 2012

This study identified motives for writing apparel product reviews in online stores, and determined what motives increase the behavior of writing reviews. It also classified store customers based on the type of writing motives, and clarified the characteristics of internet purchase behavior and of a demographic profile. Data were collected from 252 females aged 20s' and 30s' who have experience of reading and writing reviews on online shopping. The five types of writing motives were altruistic information sharing, remedying of a grievance and vengeance, economic incentives, helping new product development, and the expression of satisfaction feelings. Among five motives, altruistic information sharing, economic incentives, and helping new product development stimulate writing reviews. Store customers who write reviews were classified into three groups based on their writing motive types: Other consumer advocates(29.8%), self-interested shoppers(40.5%) and shoppers with moderate motives(29.8%). There were significant differences among three groups in writing behavior (the frequency of writing reviews, writing intent of reviews, duration of writing reviews, and frequency of online shopping) and age. Based on results, managerial implications were suggested. Long Abstract : The purpose of present study is to identify the types of writing motives on online shopping, and to clarify the motives affecting the behavior of writing reviews. This study also classifies online shoppers based on the motive types, and identifies the characteristics of the classified groups in terms of writing behavior, frequency of online shopping, and demographics. Use and Gratification Theory was adopted in this study. Qualitative research (focus group interview) and quantitative research were used. Korean women(20 to 39 years old) who reported experience with purchasing clothing online, and reading and writing reviews were selected as samples(n=252). Most of the respondents were relatively young (20-34yrs., 86.1%,), single (61.1%), employed(61.1%) and residents living in big cities(50.9%). About 69.8% of respondents read and 40.5% write apparel reviews frequently or very frequently. 24.6% of the respondents indicated an "average" in their writing frequency. Based on the qualitative result of focus group interviews and previous studies on motives for online community activities, measurement items of motives for writing after-purchase reviews were developed. All items were used a five-point Likert scale with endpoints 1 (strongly disagree) and 5 (strongly agree). The degree of writing behavior was measured by items concerning experience of writing reviews, frequency of writing reviews, amount of writing reviews, and intention of writing reviews. A five-point scale(strongly disagree-strongly agree) was employed. SPSS 18.0 was used for exploratory factor analysis, K-means cluster analysis, one-way ANOVA(Scheffe test) and ${\chi}^2$-test. Confirmatory factor analysis and path model analysis were conducted by AMOS 18.0. By conducting principal components factor analysis (varimax rotation, extracting factors with eigenvalues above 1.0) on the measurement items, five factors were identified: Altruistic information sharing, remedying of a grievance and vengeance, economic incentives, helping new product development, and expression of satisfaction feelings(see Table 1). The measurement model including these final items was analyzed by confirmatory factor analysis. The measurement model had good fit indices(GFI=.918, AGFI=.884, RMR=.070, RMSEA=.054, TLI=.941) except for the probability value associated with the ${\chi}^2$ test(${\chi}^2$=189.078, df=109, p=.00). Convergent validities of all variables were confirmed using composite reliability. All SMC values were found to be lower than AVEs confirming discriminant validity. The path model's goodness-of-fit was greater than the recommended limits based on several indices(GFI=.905, AGFI=.872, RMR=.070, RMSEA=.052, TLI=.935; ${\chi}^2$=260.433, df=155, p=.00). Table 2 shows that motives of altruistic information sharing, economic incentives and helping new product development significantly increased the degree of writing product reviews of online shopping. In particular, the effect of altruistic information sharing and pursuit of economic incentives on the behavior of writing reviews were larger than the effect of helping new product development. As shown in table 3, online store shoppers were classified into three groups: Other consumer advocates (29.8%), self-interested shoppers (40.5%), and moderate shoppers (29.8%). There were significant differences among the three groups in the degree of writing reviews (experience of writing reviews, frequency of writing reviews, amount of writing reviews, intention of writing reviews, and duration of writing reviews, frequency of online shopping) and age. For five aspects of writing behavior, the group of other consumer advocates who is mainly comprised of 20s had higher scores than the other two groups. There were not any significant differences between self-interested group and moderate group regarding writing behavior and demographics.

• Journal of Environmental Science International
• /
• v.16 no.4
• /
• pp.523-532
• /
• 2007

This study utilized the 1/25,000 topographic map of the upper area from the Geum-ho watermark located at the middle of Geum-ho river from the National Geographic Information Institute. For the analysis, first, the influence of the size of critical area to the hydro topographic factors was examined changing grid size to $10m{\times}10m,\;30m{\times}30m\;and\;50m{\times}50m$, and the critical area for the formation of a river to $0.01km^2{\sim}0.50km^2$. It is known from the examination result of watershed morphology according to the grid size that the smaller grid size, the better resolution and accuracy. And it is found, from the analysis result of the degree of the river according to the minimum critical area for each grid size, that the grid size does not affect on the degree of the river, and the number of rivers with 2nd and higher degree does not show remarkable difference while there is big difference in the number of 1st degree rivers. From the results above, it is thought that the critical area of $0.15km^2{\sim}0.20km^2$ is appropriate for formation of a river being irrelevant to the grid size in extraction of hydro topographic parameters that are used in the runoff analysis model using topographic maps. Therefore, the GIUH model applied analysis results by use of the river level difference law proposed in this study for the explanation on the outflow response-changing characters according to the decision of a critical value of a minimum level difference river, showed that, since an ogival occurrence time and an ogival flow volume are very significant in a flood occurrence in case of not undertow facilities, the researcher could obtain a good result for the forecast of river outflow when considering a convenient application of the model and an easy acquisition of data, so it's judged that this model is proper as an algorism for the decision of a critical value of a river basin.

• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.1-13
• /
• 2015

As opinion mining in big data applications has been highlighted, a lot of research on unstructured data has made. Lots of social media on the Internet generate unstructured or semi-structured data every second and they are often made by natural or human languages we use in daily life. Many words in human languages have multiple meanings or senses. In this result, it is very difficult for computers to extract useful information from these datasets. Traditional web search engines are usually based on keyword search, resulting in incorrect search results which are far from users' intentions. Even though a lot of progress in enhancing the performance of search engines has made over the last years in order to provide users with appropriate results, there is still so much to improve it. Word sense disambiguation can play a very important role in dealing with natural language processing and is considered as one of the most difficult problems in this area. Major approaches to word sense disambiguation can be classified as knowledge-base, supervised corpus-based, and unsupervised corpus-based approaches. This paper presents a method which automatically generates a corpus for word sense disambiguation by taking advantage of examples in existing dictionaries and avoids expensive sense tagging processes. It experiments the effectiveness of the method based on Naïve Bayes Model, which is one of supervised learning algorithms, by using Korean standard unabridged dictionary and Sejong Corpus. Korean standard unabridged dictionary has approximately 57,000 sentences. Sejong Corpus has about 790,000 sentences tagged with part-of-speech and senses all together. For the experiment of this study, Korean standard unabridged dictionary and Sejong Corpus were experimented as a combination and separate entities using cross validation. Only nouns, target subjects in word sense disambiguation, were selected. 93,522 word senses among 265,655 nouns and 56,914 sentences from related proverbs and examples were additionally combined in the corpus. Sejong Corpus was easily merged with Korean standard unabridged dictionary because Sejong Corpus was tagged based on sense indices defined by Korean standard unabridged dictionary. Sense vectors were formed after the merged corpus was created. Terms used in creating sense vectors were added in the named entity dictionary of Korean morphological analyzer. By using the extended named entity dictionary, term vectors were extracted from the input sentences and then term vectors for the sentences were created. Given the extracted term vector and the sense vector model made during the pre-processing stage, the sense-tagged terms were determined by the vector space model based word sense disambiguation. In addition, this study shows the effectiveness of merged corpus from examples in Korean standard unabridged dictionary and Sejong Corpus. The experiment shows the better results in precision and recall are found with the merged corpus. This study suggests it can practically enhance the performance of internet search engines and help us to understand more accurate meaning of a sentence in natural language processing pertinent to search engines, opinion mining, and text mining. Naïve Bayes classifier used in this study represents a supervised learning algorithm and uses Bayes theorem. Naïve Bayes classifier has an assumption that all senses are independent. Even though the assumption of Naïve Bayes classifier is not realistic and ignores the correlation between attributes, Naïve Bayes classifier is widely used because of its simplicity and in practice it is known to be very effective in many applications such as text classification and medical diagnosis. However, further research need to be carried out to consider all possible combinations and/or partial combinations of all senses in a sentence. Also, the effectiveness of word sense disambiguation may be improved if rhetorical structures or morphological dependencies between words are analyzed through syntactic analysis.

• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.59-83
• /
• 2018

With the increasing importance of sentiment analysis to grasp the needs of customers and the public, various types of deep learning models have been actively applied to English texts. In the sentiment analysis of English texts by deep learning, natural language sentences included in training and test datasets are usually converted into sequences of word vectors before being entered into the deep learning models. In this case, word vectors generally refer to vector representations of words obtained through splitting a sentence by space characters. There are several ways to derive word vectors, one of which is Word2Vec used for producing the 300 dimensional Google word vectors from about 100 billion words of Google News data. They have been widely used in the studies of sentiment analysis of reviews from various fields such as restaurants, movies, laptops, cameras, etc. Unlike English, morpheme plays an essential role in sentiment analysis and sentence structure analysis in Korean, which is a typical agglutinative language with developed postpositions and endings. A morpheme can be defined as the smallest meaningful unit of a language, and a word consists of one or more morphemes. For example, for a word '예쁘고', the morphemes are '예쁘(= adjective)' and '고(=connective ending)'. Reflecting the significance of Korean morphemes, it seems reasonable to adopt the morphemes as a basic unit in Korean sentiment analysis. Therefore, in this study, we use 'morpheme vector' as an input to a deep learning model rather than 'word vector' which is mainly used in English text. The morpheme vector refers to a vector representation for the morpheme and can be derived by applying an existent word vector derivation mechanism to the sentences divided into constituent morphemes. By the way, here come some questions as follows. What is the desirable range of POS(Part-Of-Speech) tags when deriving morpheme vectors for improving the classification accuracy of a deep learning model? Is it proper to apply a typical word vector model which primarily relies on the form of words to Korean with a high homonym ratio? Will the text preprocessing such as correcting spelling or spacing errors affect the classification accuracy, especially when drawing morpheme vectors from Korean product reviews with a lot of grammatical mistakes and variations? We seek to find empirical answers to these fundamental issues, which may be encountered first when applying various deep learning models to Korean texts. As a starting point, we summarized these issues as three central research questions as follows. First, which is better effective, to use morpheme vectors from grammatically correct texts of other domain than the analysis target, or to use morpheme vectors from considerably ungrammatical texts of the same domain, as the initial input of a deep learning model? Second, what is an appropriate morpheme vector derivation method for Korean regarding the range of POS tags, homonym, text preprocessing, minimum frequency? Third, can we get a satisfactory level of classification accuracy when applying deep learning to Korean sentiment analysis? As an approach to these research questions, we generate various types of morpheme vectors reflecting the research questions and then compare the classification accuracy through a non-static CNN(Convolutional Neural Network) model taking in the morpheme vectors. As for training and test datasets, Naver Shopping's 17,260 cosmetics product reviews are used. To derive morpheme vectors, we use data from the same domain as the target one and data from other domain; Naver shopping's about 2 million cosmetics product reviews and 520,000 Naver News data arguably corresponding to Google's News data. The six primary sets of morpheme vectors constructed in this study differ in terms of the following three criteria. First, they come from two types of data source; Naver news of high grammatical correctness and Naver shopping's cosmetics product reviews of low grammatical correctness. Second, they are distinguished in the degree of data preprocessing, namely, only splitting sentences or up to additional spelling and spacing corrections after sentence separation. Third, they vary concerning the form of input fed into a word vector model; whether the morphemes themselves are entered into a word vector model or with their POS tags attached. The morpheme vectors further vary depending on the consideration range of POS tags, the minimum frequency of morphemes included, and the random initialization range. All morpheme vectors are derived through CBOW(Continuous Bag-Of-Words) model with the context window 5 and the vector dimension 300. It seems that utilizing the same domain text even with a lower degree of grammatical correctness, performing spelling and spacing corrections as well as sentence splitting, and incorporating morphemes of any POS tags including incomprehensible category lead to the better classification accuracy. The POS tag attachment, which is devised for the high proportion of homonyms in Korean, and the minimum frequency standard for the morpheme to be included seem not to have any definite influence on the classification accuracy.