• Information Systems Review
• /
• v.25 no.4
• /
• pp.89-112
• /
• 2023

In March 2020, as it was declared a COVID-19 pandemic, various quarantine measures were taken. Accordingly, many changes have occurred in the tourism and hospitality industries. In particular, quarantine guidelines, such as the introduction of non-face-to-face services and social distancing, were implemented in the restaurant industry. For decades, research on restaurant attributes has emphasized the importance of three attributes: atmosphere, service quality, and food quality. Nevertheless, to the best of our knowledge, research on restaurant attributes considering the COVID-19 situation is insufficient. To respond to this call, this study attempted an exploratory approach to classify new restaurant attributes based on understanding environmental changes. This study considered 31,115 online reviews registered in Naverplace as an analysis unit, with 475 general restaurants located in Euljiro, Seoul. Further, we attempted to classify restaurant attributes by clustering words within online reviews through TF-IDF and LDA topic modeling techniques. As a result of the analysis, the factors of "prevention of infectious diseases" were derived as new attributes of restaurants in the context of COVID-19 situations, along with the atmosphere, service quality, and food quality. This study is of academic significance by expanding the literature of existing restaurant attributes in that it categorized the three attributes presented by existing restaurant attributes and further presented new attributes. Moreover, the analysis results have led to the formulation of practical recommendations, considering both the operational aspects of restaurants and policy implications.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.19 no.3
• /
• pp.109-121
• /
• 2024

This paper adopts a resource-based approach to analyze why some universities have a greater number of faculty startups, and how this impacts on performance, in terms of indictors such as the number of employees and revenue sales. More specifically, we propose 9 hypotheses which link institutional resources to faculty startups and their performance, and compare 5 different groups of university resources for cross-college variation, using data from 134 South Korean four-year universities from 2017 to 2020. We find that the institutional factors impacting on performance of faculty startups differ from other categories of startups. The results show that it is important for universities to provide a more favorable environment, incorporating more flexible personnel policies and accompanying startup support infrastructure, for faculty startups, whilest it is more effective to have more financial resources and intellectual property for other categories of startups. Our findings also indicate that university technology-holding company and technology transfer programs are crucial to increase the number of faculty startups and their performance. Our analysis results have implications for both university and government policy-makers, endeavoring to facilitate higher particaption of professors in startup formation and ultimate commercialization of associated teachnologies.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.139-156
• /
• 2021

The biggest reason for using a deep learning model in image classification is that it is possible to consider the relationship between each region by extracting each region's features from the overall information of the image. However, the CNN model may not be suitable for emotional image data without the image's regional features. To solve the difficulty of classifying emotion images, many researchers each year propose a CNN-based architecture suitable for emotion images. Studies on the relationship between color and human emotion were also conducted, and results were derived that different emotions are induced according to color. In studies using deep learning, there have been studies that apply color information to image subtraction classification. The case where the image's color information is additionally used than the case where the classification model is trained with only the image improves the accuracy of classifying image emotions. This study proposes two ways to increase the accuracy by incorporating the result value after the model classifies an image's emotion. Both methods improve accuracy by modifying the result value based on statistics using the color of the picture. When performing the test by finding the two-color combinations most distributed for all training data, the two-color combinations most distributed for each test data image were found. The result values were corrected according to the color combination distribution. This method weights the result value obtained after the model classifies an image's emotion by creating an expression based on the log function and the exponential function. Emotion6, classified into six emotions, and Artphoto classified into eight categories were used for the image data. Densenet169, Mnasnet, Resnet101, Resnet152, and Vgg19 architectures were used for the CNN model, and the performance evaluation was compared before and after applying the two-stage learning to the CNN model. Inspired by color psychology, which deals with the relationship between colors and emotions, when creating a model that classifies an image's sentiment, we studied how to improve accuracy by modifying the result values based on color. Sixteen colors were used: red, orange, yellow, green, blue, indigo, purple, turquoise, pink, magenta, brown, gray, silver, gold, white, and black. It has meaning. Using Scikit-learn's Clustering, the seven colors that are primarily distributed in the image are checked. Then, the RGB coordinate values of the colors from the image are compared with the RGB coordinate values of the 16 colors presented in the above data. That is, it was converted to the closest color. Suppose three or more color combinations are selected. In that case, too many color combinations occur, resulting in a problem in which the distribution is scattered, so a situation fewer influences the result value. Therefore, to solve this problem, two-color combinations were found and weighted to the model. Before training, the most distributed color combinations were found for all training data images. The distribution of color combinations for each class was stored in a Python dictionary format to be used during testing. During the test, the two-color combinations that are most distributed for each test data image are found. After that, we checked how the color combinations were distributed in the training data and corrected the result. We devised several equations to weight the result value from the model based on the extracted color as described above. The data set was randomly divided by 80:20, and the model was verified using 20% of the data as a test set. After splitting the remaining 80% of the data into five divisions to perform 5-fold cross-validation, the model was trained five times using different verification datasets. Finally, the performance was checked using the test dataset that was previously separated. Adam was used as the activation function, and the learning rate was set to 0.01. The training was performed as much as 20 epochs, and if the validation loss value did not decrease during five epochs of learning, the experiment was stopped. Early tapping was set to load the model with the best validation loss value. The classification accuracy was better when the extracted information using color properties was used together than the case using only the CNN architecture.

• Radiation Oncology Journal
• /
• v.20 no.1
• /
• pp.1-16
• /
• 2002

Purpose : Three dimensional conformal radiotherapy planning is being used widely for the treatment of patients with brain tumor. However, it takes much time to develop an optimal treatment plan, therefore, it is difficult to apply this technique to all patients. To increase the efficiency of this technique, we need to develop standard radiotherapy plant for each site of the brain. Therefore we developed several 3 dimensional conformal radiotherapy plans (3D plans) for tumors at each site of brain, compared them with each other, and with 2 dimensional radiotherapy plans. Finally model plans for each site of the brain were decide. Materials and Methods : Imaginary tumors, with sizes commonly observed in the clinic, were designed for each site of the brain and drawn on CT images. The planning target volumes (PTVs) were as follows; temporal $tumor-5.7\times8.2\times7.6\;cm$, suprasellar $tumor-3\times4\times4.1\;cm$, thalamic $tumor-3.1\times5.9\times3.7\;cm$, frontoparietal $tumor-5.5\times7\times5.5\;cm$, and occipitoparietal $tumor-5\times5.5\times5\;cm$. Plans using paralled opposed 2 portals and/or 3 portals including fronto-vertex and 2 lateral fields were developed manually as the conventional 2D plans, and 3D noncoplanar conformal plans were developed using beam's eye view and the automatic block drawing tool. Total tumor dose was 54 Gy for a suprasellar tumor, 59.4 Gy and 72 Gy for the other tumors. All dose plans (including 2D plans) were calculated using 3D plan software. Developed plans were compared with each other using dose-volume histograms (DVH), normal tissue complication probabilities (NTCP) and variable dose statistic values (minimum, maximum and mean dose, D5, V83, V85 and V95). Finally a best radiotherapy plan for each site of brain was selected. Results : 1) Temporal tumor; NTCPs and DVHs of the normal tissue of all 3D plans were superior to 2D plans and this trend was more definite when total dose was escalated to 72 Gy (NTCPs of normal brain 2D $plans:27\%,\;8\%\rightarrow\;3D\;plans:1\%,\;1\%$). Various dose statistic values did not show any consistent trend. A 3D plan using 3 noncoplanar portals was selected as a model radiotherapy plan. 2) Suprasellar tumor; NTCPs of all 3D plans and 2D plans did not show significant difference because the total dose of this tumor was only 54 Gy. DVHs of normal brain and brainstem were significantly different for different plans. D5, V85, V95 and mean values showed some consistent trend that was compatible with DVH. All 3D plans were superior to 2D plans even when 3 portals (fronto-vertex and 2 lateral fields) were used for 2D plans. A 3D plan using 7 portals was worse than plans using fewer portals. A 3D plan using 5 noncoplanar portals was selected as a model plan. 3) Thalamic tumor; NTCPs of all 3D plans were lower than the 2D plans when the total dose was elevated to 72 Gy. DVHs of normal tissues showed similar results. V83, V85, V95 showed some consistent differences between plans but not between 3D plans. 3D plans using 5 noncoplanar portals were selected as a model plan. 4) Parietal (fronto- and occipito-) tumors; all NTCPs of the normal brain in 3D plans were lower than in 2D plans. DVH also showed the same results. V83, V85, V95 showed consistent trends with NTCP and DVH. 3D plans using 5 portals for frontoparietal tumor and 6 portals for occipitoparietal tumor were selected as model plans. Conclusion : NTCP and DVH showed reasonable differences between plans and were through to be useful for comparing plans. All 3D plans were superior to 2D plans. Best 3D plans were selected for tumors in each site of brain using NTCP, DVH and finally by the planner's decision.

• Journal of Korean Society of Forest Science
• /
• v.99 no.1
• /
• pp.9-15
• /
• 2010

This study was conducted to develop a stand growth model and a stand yield table for Eucalyptus pellita and Acacia mangium plantations in Kalimantan, Indonesia. To develop a stand growth model, Weibull robability density function, a diameter class model, was applied in this study. In the development of stand growth model by site index and stand age, a hierarchy is generally required - estimation, recovery and prediction of the diameter class model. A number of grow equations were also involved in each process to estimate diameter, height, basal area, minimum or maximum diameter. To examine whether the grow equations are adequate for Eucalyptus pellita or Acacia mangium plantations, a fitness index was analyzed for each equation. The results showed that fitness indices were ranged from 65 to 89% for Eucalyptus pellita plantations and from 72 to 95% for Acacia mangium plantations. As being highly adequate for the plantations, a stand yield table was developed based on the resulted growth model, and applied to estimate the stand growth with midium site index for 10-year period. The highest annual stand growth of Eucalyptus pellita plantations was estimated to be 21.25 $m^3$/ha, while that of Acacia mangium plantations was 27.5 $m^3$/ha. In terms of annual stand growth, Acacia mangium plantations appeared to be more beneficial than Eucalyptus pellita plantations. Also, to estimate commercial timber volume available from the plantations, an assumption that a log would be cut by 2.7 m in length and the rest of the log would be cut by 1.5m was involved. The commercial timber volume available from Eucalyptus pellita plantations was 68.0 $m^3$/ha, 33% from the total stand volume, 203.2 $m^3$/ha. Also 96.7 $m^3$/ha of commercial timbers were available from Acacia mangium plantations, which was 42% from the 232.9 $m^3$/ha in total. Presenting a good information about the stand growth in Eucalyptus pellita and Acacia mangium plantations, this study might be useful for whom proceeds or considers an abroad plantation for merchantable timber production or carbon credit in tropical regions.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.2 no.4
• /
• pp.71-102
• /
• 2007

There were many comprehensive analyses conducted within the existing research activities wherein factors affecting technology progress including investment in R&D vis-${\Box}$-vis their influences act as the determinants of TFP. Note, however, that there were few comprehensive analysis in the industrial research performed regarding the impact of the economy of scale as it affects TFP; most of these research studies dealt with the analysis of the non -parametric Malmquist productivity index or used the stochastic frontier production function models. No comprehensive analysis on the impacts of individual independent variables affecting TFP was performed. Therefore, this study obtained the TFP increase rate of each industry by analyzing the factors of the existing growth accounting equation and comprehensively analyzed the TFP determinants by constructing a comprehensive analysis model considering the investment in R&D and economy of scale (smallness by industry) as the influencers of TFP by industry. First, for the TFP increase rate of the 15 industries as a whole, the annual average increase rate for 1993${\sim}$ 1997 was approximately 3.8% only; during 1999${\sim}$ 2000 following the foreign exchange crisis, however, the annual increase rate rose to approximately 7.8%. By industry, the annual average increase rate of TFP between 1993 and 2000 stood at 11.6%, the highest in the electrical and electronic equipment manufacturing business and IT manufacturing sector. In contrast, a -0.4% increase rate was recorded in the furniture and other product manufacturing sectors. In the case of the service industry, the TFP increase rate was 7.3% in the transportation, warehousing, and communication sectors. This is much higher than the 2.9% posted in the electricity, water, and gas sectors and -3.7% recorded in the wholesale, food, and hotel businesses. The results of the comprehensive analysis conducted on the determinants of TFP showed that the correlations between R&D and TFP in general were positive (+) correlations whose significance has yet to be validated; in the model where the self-employed and unpaid family workers were used as proxy variables indicating the smallness of industry out of the total number of workers, however, significant negative (-) correlations were noted. On the other hand, the estimation factors of variables surrogating the smallness of scale in each industry showed that a consistently high "smallness of scale" in an industry means a decrease in the increase rate of TFP in the same industry.

• The Journal of Korean Society for Radiation Therapy
• /
• v.29 no.2
• /
• pp.27-32
• /
• 2017

Objectives: Oncological hyperthermia is a treatment to selectively kill cancer cells by directly applying heat to cancer cells or indirectly demage cancer cells. One of the most side effects of treatment is burn that can appear on the skin. In areas with irregularities such as the umbilicus, the patient feels a sense of hot and treatment may be discontinued. Therefore, in order to eliminate the irregularities of these areas, compensators are manufactured and measured to decrease in temperature. Materials and Methods: The temperature of the four sites (umbilicus, near the umbilicus, 5 cm below the umbilicus, back) was measured five times around the umbilicus in patients who were treated at oncological hyperthermia treatment device(EHY-2000, Oncotherm Kft, Hungary). The temperature sensor (TM-100, Oncotherm Kft, Hungary) was attached to four sites and the changes were observed at 5, 15, 25, 35, and 50 minutes after treatment. Compensators of three materials were used(Vaseline, Bolus, Dental resin). The data measured five times were compared for each compensator. Results: The temperature change when the compensator was not used increase from 34.65 degrees to 42.9 degrees on average. The near umbilicus was changed from 32.20 degrees to 37.00 degrees, and the 5 cm below the umbilicus was changed from 31.90 to 34.41 degrees. When the compensator material was inserted into the umbilicus, the temperature change was measured as 5.42 degrees for bolus, 6.55 degrees for vaseline, and 6.83 degrees for resin. Conclusion: Using the compensator in the region where the irregularities such as the umbilicus, the heat sensation could be reduced. the use of a resin that can be customized not only lowers the temperature but also significantly reduces the feeling of the patient. It will be possible to reduce the heat sensation in the treatment and to treat it in a more comfortable condition.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.43 no.4
• /
• pp.570-576
• /
• 2014

Response surface methodology (RSM) was applied to optimize the microwave-assisted extraction (MAE) conditions for electron-donating ability, total phenol content, and total flavonoid content of Ligustrum lucidum Aiton. Ligustrum lucidum Aiton from different regions was tested, and Ligustrum lucidum Aiton from Haenam was chosen due to its higher total phenolic content, total flavonoid content, DPPH radical scavenging activity and ABTS radical scavenging activity compared to the other samples. Central composite design was used to optimize extraction of Ligustrum lucidum Aiton from Haenam as well as determine the effects of extraction temperature ($X_1$) and extraction time ($X_2$) on dependent variables ($Y_n$). Determination coefficients ($R^2$) of the regression equations for dependent variables ranged from 0.8858 to 0.9517. The optimum points were $131.68^{\circ}C$ for extraction temperature and 5.49 min for extraction time. Predicted values of the optimized conditions were acceptable when compared to experimental values.

• Economic and Environmental Geology
• /
• v.45 no.2
• /
• pp.71-78
• /
• 2012

Illite-smectite mixed layers (I-S) occurring authigenically in diagenetic and hydrothermal environments reacts toward more illite-rich phases as temperature and potassium ion concentration increase. For that reason, I-S is often used as geothermometry and/or geochronometry at the field of hydrocarbons or ore minerals exploration. Generally, I-S shows X-ray powder diffraction (XRD) patterns of ultra-thin lamellar structures, which consist of restricted numbers of sillicate layers (normally, 5 ~ 15 layers) stacked in parallel to a-b planes. This ultra-thinness is known to decrease I-S expandability (%S) rather than theoretically expected one (short-stacking effect). We attempt here to quantify the short stacking effect of I-S using the difference of two types of expandability: one type is a maximum expandability ($%S_{Max}$) of infinite stacks of fundamental particles (physically inseparable smallest units), and the other type is an expandability of finite particle stacks normally measured using X-ray powder diffraction (XRD) ($%S_{XRD}$). Eleven I-S samples from the Geumseongsan volcanic complex, Uiseong, Gyeongbuk, have been analyzed for measuring $%S_{XRD}$ and average coherent scattering thickness (CST) after size separation under 1 ${\mu}m$. Average fundamental particle thickness ($N_f$) and $%S_{Max}$ have been determined from $%S_{XRD}$ and CST using inter-parameter relationships of I-S layer structures. The discrepancy between $%S_{Max}$ and $%S_{XRD}$ (${\Delta}%S$) suggests that the maximum short-stacking effect happens approximately at 20 $%S_{XRD}$, of which point represents I-S layer structures consisting of ca. average 3-layered fundamental particles ($N_f{\approx}3$). As a result of inferring the $%S_{XRD}$ range of each Reichweite using the $%S_{XRD}$ vs. $N_f$ diagram of Kang et al. (2002), we can confirms that the fundamental particle thickness is a determinant factor for I-S Reichweite, and also that the short-stacking effect shifts the $%S_{XRD}$ range of each Reichweite toward smaller $%S_{XRD}$ values than those that can be theoretically prospected using junction probability.

• The KIPS Transactions:PartB
• /
• v.11B no.6
• /
• pp.749-758
• /
• 2004

In this paper, we propose a new method utilizing only raw corpus without additional human effort for disambiguation of target word selection in English-Korean machine translation. We use two data-driven techniques; one is the Latent Semantic Analysis(LSA) and the other the Probabilistic Latent Semantic Analysis(PLSA). These two techniques can represent complex semantic structures in given contexts like text passages. We construct linguistic semantic knowledge by using the two techniques and use the knowledge for target word selection in English-Korean machine translation. For target word selection, we utilize a grammatical relationship stored in a dictionary. We use k- nearest neighbor learning algorithm for the resolution of data sparseness Problem in target word selection and estimate the distance between instances based on these models. In experiments, we use TREC data of AP news for construction of latent semantic space and Wail Street Journal corpus for evaluation of target word selection. Through the Latent Semantic Analysis methods, the accuracy of target word selection has improved over 10% and PLSA has showed better accuracy than LSA method. finally we have showed the relatedness between the accuracy and two important factors ; one is dimensionality of latent space and k value of k-NT learning by using correlation calculation.