• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.43-56
• /
• 2020

Recommender systems help users make the best choice among various options. Especially, recommender systems play important roles in internet sites as digital information is generated innumerable every second. Many studies on recommender systems have focused on an accurate recommendation. However, there are some problems to overcome in order for the recommendation system to be commercially successful. First, there is a lack of transparency in the recommender system. That is, users cannot know why products are recommended. Second, the recommender system cannot immediately reflect changes in user preferences. That is, although the preference of the user's product changes over time, the recommender system must rebuild the model to reflect the user's preference. Therefore, in this study, we proposed a recommendation methodology using topic modeling and sequential association rule mining to solve these problems from review data. Product reviews provide useful information for recommendations because product reviews include not only rating of the product but also various contents such as user experiences and emotional state. So, reviews imply user preference for the product. So, topic modeling is useful for explaining why items are recommended to users. In addition, sequential association rule mining is useful for identifying changes in user preferences. The proposed methodology is largely divided into two phases. The first phase is to create user profile based on topic modeling. After extracting topics from user reviews on products, user profile on topics is created. The second phase is to recommend products using sequential rules that appear in buying behaviors of users as time passes. The buying behaviors are derived from a change in the topic of each user. A collaborative filtering-based recommendation system was developed as a benchmark system, and we compared the performance of the proposed methodology with that of the collaborative filtering-based recommendation system using Amazon's review dataset. As evaluation metrics, accuracy, recall, precision, and F1 were used. For topic modeling, collapsed Gibbs sampling was conducted. And we extracted 15 topics. Looking at the main topics, topic 1, top 3, topic 4, topic 7, topic 9, topic 13, topic 14 are related to "comedy shows", "high-teen drama series", "crime investigation drama", "horror theme", "British drama", "medical drama", "science fiction drama", respectively. As a result of comparative analysis, the proposed methodology outperformed the collaborative filtering-based recommendation system. From the results, we found that the time just prior to the recommendation was very important for inferring changes in user preference. Therefore, the proposed methodology not only can secure the transparency of the recommender system but also can reflect the user's preferences that change over time. However, the proposed methodology has some limitations. The proposed methodology cannot recommend product elaborately if the number of products included in the topic is large. In addition, the number of sequential patterns is small because the number of topics is too small. Therefore, future research needs to consider these limitations.

• The KIPS Transactions:PartB
• /
• v.13B no.5 s.108
• /
• pp.561-568
• /
• 2006

WPS(Wearable Personal Station) for next generation PC can define as a core terminal of 'Ubiquitous Computing' that include information processing and network function and overcome spatial limitation in acquisition of new information. As a way to acquire significant dynamic gesture data of user from haptic devices, traditional gesture recognizer based on desktop-PC using wire communication module has several restrictions such as conditionality on space, complexity between transmission mediums(cable elements), limitation of motion and incommodiousness on use. Accordingly, in this paper, in order to overcome these problems, we implement hand gesture recognition system using fuzzy algorithm and neural network for Post PC(the embedded-ubiquitous environment using blue-tooth module and WPS). Also, we propose most efficient and reasonable hand gesture recognition interface for Post PC through evaluation and analysis of performance about each gesture recognition system. The proposed gesture recognition system consists of three modules: 1) gesture input module that processes motion of dynamic hand to input data 2) Relational Database Management System(hereafter, RDBMS) module to segment significant gestures from input data and 3) 2 each different recognition modulo: fuzzy max-min and neural network recognition module to recognize significant gesture of continuous / dynamic gestures. Experimental result shows the average recognition rate of 98.8% in fuzzy min-nin module and 96.7% in neural network recognition module about significantly dynamic gestures.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.1
• /
• pp.431-438
• /
• 2021

To prepare more stringent regulations for USCG Phase II ballast water management, this study investigated the staining efficiency of SYBR Green I(SGI) and SYBR Gold(SG) on the virus-like particle (VLP). A dye with high staining efficiency was applied to the treated water that was passed through the ballast water treatment system (BWTS). VLP staining was observed most clearly under the 100-fold and 200-fold dilution of the stock solution when the volume of filtered samples was 0.5 mL to 2 mL. The staining efficiency of SGI and SG did not show a significant difference. On the other hand, the green fluorescence of viruses in the sample stained with SGI was more pronounced than in the samples stained with SG (expressed yellow fluorescence), making it easier to observe. The abundance of VLP in the test water and control water treatments that did not pass through the two types of BWTS (electrolysis type, UV + electrolysis type) was approximately 109 - 1010 VLP 100 mL-1. In contrast, no stained VLP was observed in the treated water treatments. Moreover, SGI was confirmed to be effectively stained under various salinity conditions, including seawater, brackish water, and freshwater. Further verification tests and development of staining methods under various BWTS are required, but the SGI staining method is believed to be a good alternative to the VLP live/dead determination of the USCG Phase II type approval test.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.15 no.6
• /
• pp.27-42
• /
• 2020

Startup accelerators have emerged as new investment entities that help early startups, which are not easy to survive continuously due to lack of funds, commercialization capabilities, and experiences. As their positive performance on early startups and the ecosystem has been proven, the number of early startups which want to receive their investment is also increasing. However, they are vaguely preparing to attract accelerators' investment because they do not have any information on what factors the accelerators consider important. In addition, researches on startup accelerators are also at an early level, so there are no remarkable prior studies on factors that decide on investment. Therefore, this study aims to help startups prepare for investment attraction by looking at what factors are important for accelerators to invest, and to provide meaningful implications to academia. In the preceding study, we derived five upper level categories, 26 lower level accelerators' investment determinants through the qualitative meta-synthesis method, secondary data analysis, observation on US accelerators and in-depth interviews. In this study, we want to derive important implications by deriving priorities of the accelerators' investment determinants. Therefore, we used AHP that are evaluated as the suitable methodology for deriving importance and priority. The analysis results show that accelerators value market-related factors most. This means that startups that are subject to investment by accelerators are early-stage startups, and many companies have not fully developed their products or services. Therefore, market-related factors that can be evaluated objectively seem to be more important than products (or services) that are still ambiguous. Next, it was found that the factors related to the internal workforce of startups are more important. Since accelerators want to develop their businesses together with start-ups and team members through mentoring, ease of collaboration with them is very important, which seems to be important. The overall priority analysis results of the 26 investment determinants show that 'customer needs' and 'founders and team members' understanding of customers and markets' (0.62) are important and high priority factors. The results also show that startup accelerators consider the customer-centered perspective very important. And among the factors related to startups, the most prominent factor was the founder's openness and execution ability. Therefore, it can be confirmed that accelerators consider the ease of collaboration with these startups very important.

• Journal of Bio-Environment Control
• /
• v.31 no.1
• /
• pp.43-51
• /
• 2022

This study was conducted to investigate the effect of ventilation at high temperature on the control of powdery mildew, silverleaf whitefly two-spotted spider mite occurred at Korean melon cultivation greenhouse, and on leaf rolling and flowering of the plant in summer season. 'Alchanggul' grafted onto 'Hidden Power' rootstock was planted on soil bed with the distance of 40 cm. Three ventilation temperatures of 45℃, 40℃, and 35℃ as set points were compared. Ventilation treatment was done by control of side window operation from 18th June to 13th July when silverleaf whitefly, mite, and powdery mildew were occurred in all greenhouses. The temperature inside greenhouse was increased up to the set temperature point on sunny days and maintained for about 9 hours with high relative humidity at 45℃ condition. The differences of day maximum air temperature and day minimum RH were the highest at 45℃ treatment. After 11 days of treatments, the damage by powdery mildew and two-spotted spider mite was almost recovered at 45℃ treatment but not at 40 and 35℃. The population of silverleaf whitefly and two-spotted spider mite were significantly decreased at 45℃ treatment at 14 days after treatment, while powdery mildew symptom was not significantly decreased. Leaf rolling was observed at high temperature but not severe at 45℃ treatment. After 26 days of treatments, female flowers did not bloom at all at 45℃ treatment, and the number of male flowers was 1.2 among 15 nodes of newly grown shoots. As the result, it indicates that ventilation at the high temperature of 45℃ for about 2 to 3 weeks can be an applicable method to control above mentioned pests and disease, and to recover the vegetative growth of Korean melon by reducing flowering of the plant.

• Journal of Bio-Environment Control
• /
• v.31 no.4
• /
• pp.504-512
• /
• 2022

To examine the distribution of internal solar radiation within various locations in multi-span greenhouses, the solar radiation, light transmittance, and accumulated radiation at the central and lateral sections were analyzed by dividing 8:30 to 12:30 in the morning and 12:35 to 16:30 in the afternoon. The growth and yield of tomatoes within these sections were also compared. In the morning, the solar radiation of the central section and the side section was 275.2 W·m^-2 and 314.9 W·m^-2, while in the afternoon, it was 314.9 W·m^-2 and 313.9 W·m^-2, respectively. The light transmittance and accumulated radiation were also low, confirming the low distribution of solar radiation in the central (connecting) section of the multi-span greenhouses. The growth survey revealed no significant difference. The final yield of tomatoes per plant was 4,828 g in the central section and 4,851 g in the lateral section, but there was no significant difference in the central section compared to the lateral section by 0.5%. However, the amount of solar radiation as per time in the central section is higher than the light compensation point, 60 W·m^-2, and slightly lower than the light saturation point of tomatoes, i.e., 281 W·m^-2. The results of this study can help in greenhouse design based on the insolation environment.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.51 no.1
• /
• pp.13-28
• /
• 2023

This study examines the process of project promotion in Busan, which introduced the private participation-type consultative body for the first time in the country in the promotion stage of the private-initiated park development project, and introduced flexible application of the system and differentiated policy elements in the implementation process, and examines the operational characteristics and we tried to analyze performance, etc. As a result of the analysis, first, the preferred bidder was selected by introducing a mixed method in the project method, which is an independent project method that cannot be seen in other local governments. Second, by specifying guidelines considering the characteristics of each park and detailed guidelines such as the location, area, and maximum height of non-park facilities, criteria for establishing a rational development plan utilizing regional identity and the basis of evaluation standards were laid. Third, in the project process, transparency was secured through the delegation-type roundtable, in which the private sector performs the functions and roles of key actors, thereby preventing disputes such as suspicion of preferential treatment. Fourth, in order to improve the quality of park facilities to be donated and to secure design adequacy, after approval of the implementation plan, a general planner was introduced and construction project management (design stage) services were performed to promote efficient implementation and specialization of luxury parks in the region. As a result, the city of Busan carried out the project efficiently by conserving 5 parks from sunset, a park area of 2.25km², and reducing land compensation and park construction costs by KRW 740 billion. Reinforcement of the public nature of the private-initiated park development project was suggested. However, due to the application of these systems and verification procedures, the project period is prolonged, and park services are delayed along with the financial burden on private operators.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_1
• /
• pp.655-667
• /
• 2023

Algal bloom outbreaks are frequently reported around the world, and serious water pollution problems arise every year in Korea. It is necessary to protect the aquatic ecosystem through continuous management and rapid response. Many studies using satellite images are being conducted to estimate the concentration of chlorophyll-a (Chl-a), an indicator of algal bloom occurrence. However, machine learning models have recently been used because it is difficult to accurately calculate Chl-a due to the spectral characteristics and atmospheric correction errors that change depending on the water system. It is necessary to consider the factors affecting algal bloom as well as the satellite spectral index. Therefore, this study constructed a dataset by considering water quality, hydrological and meteorological factors, and sentinel-2 images in combination. Representative ensemble models random forest and extreme gradient boosting (XGBoost) were used to predict the concentration of Chl-a in eight weirs located on the Nakdong river over the past five years. R-squared score (R²), root mean square errors (RMSE), and mean absolute errors (MAE) were used as model evaluation indicators, and it was confirmed that R² of XGBoost was 0.80, RMSE was 6.612, and MAE was 4.457. Shapley additive expansion analysis showed that water quality factors, suspended solids, biochemical oxygen demand, dissolved oxygen, and the band ratio using red edge bands were of high importance in both models. Various input data were confirmed to help improve model performance, and it seems that it can be applied to domestic and international algal bloom detection.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.12 no.3
• /
• pp.235-240
• /
• 2008

Purpose: At the beginning of PET/CT, Computed Tomography was mainly used only for Attenuation Correction (AC), but as the performance of the CT have been increase, it could give improved diagnostic information with Contrast Media. But it was controversial that Contrast Media could affect AC on PET/CT scan. Some submitted thesis' show that Contrast Media could overestimate when it is for AC data processing. On the contrary, the opinion that Contrast Media could be possible to affect the alteration of SUV because of the overestimated AC. But it does not have a definite effect on the diagnosis. Thus, the affection of Contrast Media on AC was investigated in this study. Materials and Methods: Patient inclusion criteria required a history of a malignancy and performance of an integrated PET/CT scan and contrast- enhanced CT scan within a 1-day period. Thirty oncologic patients who had PET/CT scan from December 2007 to June 2008 underwent staging evaluation and met these criteria. All patients fasted for at least 6 hr before the IV injection of approximately 5.6 MBq/kg (0.15 mCi/kg) of $^{18}F$-FDG and were scanned about 60 min after injection. All patients had a whole body PET/CT performed without IV contrast media followed by a contrast-enhanced CT on the Discovery STe PET/CT scanner. CT data were used for AC and PET images came out after AC. The ROIs drew and measured SUV. A paired t-test of these results was performed to assess the significance of the difference between the SUV obtained from the two attenuation corrected PET images. Results: The mean and maximum Standardized Uptake Values (SUV) for different regions averaged over all Patients. Comparing before using Contrast Media and after using, Most of ROIs have the increased SUV when it did Contrast Enhanced CT compare to Non-Contrast enhanced CT. All regions have increased SUV and also their p value was under 0.05 except the mean SUV of the Heart region. Conclusion: In this regard, the effect on SUV measurements that occurs when a contrast-enhanced CT is used for attenuation correction could have significant clinical ramifications. But some submitted thesis insisted that the percentage change in SUV that can determine or modify clinical management of oncology patients is small. Because there was not much difference that could be discovered by interpreter. But obviously the numerical change was occurred and on the stage finding primary region, small change would be base line, such as the region of liver which has greater change than the other regions needs more attention.

• 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.