• Journal of Intelligence and Information Systems
• /
• v.23 no.4
• /
• pp.1-31
• /
• 2017

Modern businesses are adopting new technologies to serve their markets better as well as to improve efficiency and productivity. The advertising industry has continuously experienced disruptions from the traditional channels (radio, television and print media) to new complex ones including internet, social media and mobile-based advertising. This case study focuses on proposing intelligent advertising business model in Seoul's metro network. Seoul has one of the world's busiest metro network and transports a huge number of travelers on a daily basis. The high number of travelers coupled with a well-planned metro network creates a platform where marketers can initiate engagement and interact with both customers and potential customers. In the current advertising model, advertising is on illuminated and framed posters in the stations and in-car, non-illuminated posters, and digital screens that show scheduled arrivals and departures of metros. Some stations have digital screens that show adverts but they do not have location capability. Most of the current advertising media have one key limitation: space. For posters whether illuminated or not, one space can host only one advert at a time. Empirical literatures show that there is room for improving this advertising model and eliminate the space limitation by replacing the poster adverts with digital advertising platform. This new model will not only be digital, but will also provide intelligent advertising platform that is driven by data. The digital platform will incorporate location sensing, e-commerce, and mobile platform to create new value to all stakeholders. Travel cards used in the metro will be registered and the card scanners will have a capability to capture traveler's data when travelers tap their cards. This data once analyzed will make it possible to identify different customer groups. Advertisers and marketers will then be able to target specific customer groups, customize adverts based on the targeted consumer group, and offer a wide variety of advertising formats. Format includes video, cinemagraphs, moving pictures, and animation. Different advert formats create different emotions in the customer's mind and the goal should be to use format or combination of formats that arouse the expected emotion and lead to an engagement. Combination of different formats will be more effective and this can only work in a digital platform. Adverts will be location based, ensuring that adverts will show more frequently when the metro is near the premises of an advertiser. The advertising platform will automatically detect the next station and screens inside the metro will prioritize adverts in the station where the metro will be stopping. In the mobile platform, customers who opt to receive notifications will receive them when they approach the business premises of advertiser. The mobile platform will have indoor navigation for the underground shopping malls that will allow customers to search for facilities within the mall, products they may want to buy as well as deals going on in the underground mall. To create an end-to-end solution, the mobile solution will have a capability to allow customers purchase products through their phones, get coupons for deals, and review products and shops where they have bought a product. The indoor navigation will host intelligent mobile-based advertisement and a recommendation system. The indoor navigation will have adverts such that when a customer is searching for information, the recommendation system shows adverts that are near the place traveler is searching or in the direction that the traveler is moving. These adverts will be linked to the e-commerce platform such that if a customer clicks on an advert, it leads them to the product description page. The whole system will have multi-language as well as text-to-speech capability such that both locals and tourists have no language barrier. The implications of implementing this model are varied including support for small and medium businesses operating in the underground malls, improved customer experience, new job opportunities, additional revenue to business model operator, and flexibility in advertising. The new value created will benefit all the stakeholders.

• Journal of the Korean Child Neurology Society
• /
• v.26 no.4
• /
• pp.205-209
• /
• 2018

Purpose: Several studies have shown that magnesium plays an important role in modulating N-methyl-D-aspartate (NMDA)-related seizures by blocking NMDA ion channel receptors. Clinicians usually measure total serum magnesium levels instead of biologically active ionized magnesium levels. We compared the serum ionized magnesium ($iMg^{2+}$) level between epileptic children with and without a history of fever-triggered seizure (FTS). Methods: All epileptic children who visited the outpatient clinic or pediatric emergency department at Korea University Guro Hospital between January 2015 and July 2017 were included. Only epileptic children aged 1-8 years who were newly diagnosed within 2 years were included. Results: There were 12 children with FTS and 16 without FTS. Median serum $iMg^{2+}$ level was 0.93 (0.85-1.14, quartile) mEq/L. Serum $iMg^{2+}$ level was significantly lower in epileptic children with FTS (0.86 mEq/L) compared to those without FTS (1.10 mEq/L) (P=0.005). No difference was noted in clinical variables between the two groups. Lower serum $iMg^{2+}$ level significantly increased the risk of having FTS in epileptic children based on multivariable logistic regression analysis (odds ratio [OR]=0.028). Conclusion: Serum $iMg^{2+}$ level was significantly lower in epileptic children with FTS than in those without FTS. Measurement of biologically active serum $iMg^{2+}$ level could be considered in epileptic children with recurrent FTS. A large-scale prospective study is warranted.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_2
• /
• pp.1295-1305
• /
• 2021

Satellite-derived ocean color products are required to effectively monitor clear open ocean and coastal water regions for various research fields. For this purpose, accurate correction of atmospheric effect is essential. Currently, the Geostationary Ocean Color Imager (GOCI)-II ground segment uses the reanalysis of meteorological fields such as European Centre for Medium-Range Weather Forecasts (ECMWF) or National Centers for Environmental Prediction (NCEP) to correct gas absorption by water vapor and ozone. In this process, uncertainties may occur due to the low spatiotemporal resolution of the meteorological data. In this study, we develop water vapor absorption correction model for the GK-2 combined GOCI-II atmospheric correction using Advanced Meteorological Imager (AMI) total precipitable water (TPW) information through radiative transfer model simulations. Also, we investigate the impact of the developed model on GOCI products. Overall, the errors with and without water vapor absorption correction in the top-of-atmosphere (TOA) reflectance at 620 nm and 680 nm are only 1.3% and 0.27%, indicating that there is no significant effect by the water vapor absorption model. However, the GK-2A combined water vapor absorption model has the large impacts at the 709 nm channel, as revealing error of 6 to 15% depending on the solar zenith angle and the TPW. We also found more significant impacts of the GK-2 combined water vapor absorption model on Rayleigh-corrected reflectance at all GOCI-II spectral bands. The errors generated from the TOA reflectance is greatly amplified, showing a large error of 1.46~4.98, 7.53~19.53, 0.25~0.64, 14.74~40.5, 8.2~18.56, 5.7~11.9% for from 620 nm to 865 nm, repectively, depending on the SZA. This study emphasizes the water vapor correction model can affect the accuracy and stability of ocean color products, and implies that the accuracy of GOCI-II ocean color products can be improved through fusion with GK-2A/AMI.

• Korean Journal of Remote Sensing
• /
• v.39 no.2
• /
• pp.207-221
• /
• 2023

This study suggests deep neural network models for estimating air temperature with Level 1B (L1B) datasets of GEO-KOMPSAT-2A (GK-2A). The temperature at 1.5 m above the ground impact not only daily life but also weather warnings such as cold and heat waves. There are many studies to assume the air temperature from the land surface temperature (LST) retrieved from satellites because the air temperature has a strong relationship with the LST. However, an algorithm of the LST, Level 2 output of GK-2A, works only clear sky pixels. To overcome the cloud effects, we apply a deep neural network (DNN) model to assume the air temperature with L1B calibrated for radiometric and geometrics from raw satellite data and compare the model with a linear regression model between LST and air temperature. The root mean square errors (RMSE) of the air temperature for model outputs are used to evaluate the model. The number of 95 in-situ air temperature data was 2,496,634 and the ratio of datasets paired with LST and L1B show 42.1% and 98.4%. The training years are 2020 and 2021 and 2022 is used to validate. The DNN model is designed with an input layer taking 16 channels and four hidden fully connected layers to assume an air temperature. As a result of the model using 16 bands of L1B, the DNN with RMSE 2.22℃ showed great performance than the baseline model with RMSE 3.55℃ on clear sky conditions and the total RMSE including overcast samples was 3.33℃. It is suggested that the DNN is able to overcome cloud effects. However, it showed different characteristics in seasonal and hourly analysis and needed to append solar information as inputs to make a general DNN model because the summer and winter seasons showed a low coefficient of determinations with high standard deviations.

• Journal of Intelligence and Information Systems
• /
• v.23 no.4
• /
• pp.77-110
• /
• 2017

Recently, the diversification and individualization of consumption patterns through the web and mobile devices based on the Internet have been rapid. As this happens, the efficient operation of the offline store, which is a traditional distribution channel, has become more important. In order to raise both the sales and profits of stores, stores need to supply and sell the most attractive products to consumers in a timely manner. However, there is a lack of research on which SKUs, out of many products, can increase sales probability and reduce inventory costs. In particular, if a company sells products through multiple in-store stores across multiple locations, it would be helpful to increase sales and profitability of stores if SKUs appealing to customers are recommended. In this study, the recommender system (recommender system such as collaborative filtering and hybrid filtering), which has been used for personalization recommendation, is suggested by SKU recommendation method of a store unit of a distribution company that handles a homogeneous brand through a plurality of sales stores by country and region. We calculated the similarity of each store by using the purchase data of each store's handling items, filtering the collaboration according to the sales history of each store by each SKU, and finally recommending the individual SKU to the store. In addition, the store is classified into four clusters through PCA (Principal Component Analysis) and cluster analysis (Clustering) using the store profile data. The recommendation system is implemented by the hybrid filtering method that applies the collaborative filtering in each cluster and measured the performance of both methods based on actual sales data. Most of the existing recommendation systems have been studied by recommending items such as movies and music to the users. In practice, industrial applications have also become popular. In the meantime, there has been little research on recommending SKUs for each store by applying these recommendation systems, which have been mainly dealt with in the field of personalization services, to the store units of distributors handling similar brands. If the recommendation method of the existing recommendation methodology was 'the individual field', this study expanded the scope of the store beyond the individual domain through a plurality of sales stores by country and region and dealt with the store unit of the distribution company handling the same brand SKU while suggesting a recommendation method. In addition, if the existing recommendation system is limited to online, it is recommended to apply the data mining technique to develop an algorithm suitable for expanding to the store area rather than expanding the utilization range offline and analyzing based on the existing individual. The significance of the results of this study is that the personalization recommendation algorithm is applied to a plurality of sales outlets handling the same brand. A meaningful result is derived and a concrete methodology that can be constructed and used as a system for actual companies is proposed. It is also meaningful that this is the first attempt to expand the research area of the academic field related to the existing recommendation system, which was focused on the personalization domain, to a sales store of a company handling the same brand. From 05 to 03 in 2014, the number of stores' sales volume of the top 100 SKUs are limited to 52 SKUs by collaborative filtering and the hybrid filtering method SKU recommended. We compared the performance of the two recommendation methods by totaling the sales results. The reason for comparing the two recommendation methods is that the recommendation method of this study is defined as the reference model in which offline collaborative filtering is applied to demonstrate higher performance than the existing recommendation method. The results of this model are compared with the Hybrid filtering method, which is a model that reflects the characteristics of the offline store view. The proposed method showed a higher performance than the existing recommendation method. The proposed method was proved by using actual sales data of large Korean apparel companies. In this study, we propose a method to extend the recommendation system of the individual level to the group level and to efficiently approach it. In addition to the theoretical framework, which is of great value.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.173-198
• /
• 2020

For a long time, many studies have been conducted on predicting the success of campaigns for customers in academia, and prediction models applying various techniques are still being studied. Recently, as campaign channels have been expanded in various ways due to the rapid revitalization of online, various types of campaigns are being carried out by companies at a level that cannot be compared to the past. However, customers tend to perceive it as spam as the fatigue of campaigns due to duplicate exposure increases. Also, from a corporate standpoint, there is a problem that the effectiveness of the campaign itself is decreasing, such as increasing the cost of investing in the campaign, which leads to the low actual campaign success rate. Accordingly, various studies are ongoing to improve the effectiveness of the campaign in practice. This campaign system has the ultimate purpose to increase the success rate of various campaigns by collecting and analyzing various data related to customers and using them for campaigns. In particular, recent attempts to make various predictions related to the response of campaigns using machine learning have been made. It is very important to select appropriate features due to the various features of campaign data. If all of the input data are used in the process of classifying a large amount of data, it takes a lot of learning time as the classification class expands, so the minimum input data set must be extracted and used from the entire data. In addition, when a trained model is generated by using too many features, prediction accuracy may be degraded due to overfitting or correlation between features. Therefore, in order to improve accuracy, a feature selection technique that removes features close to noise should be applied, and feature selection is a necessary process in order to analyze a high-dimensional data set. Among the greedy algorithms, SFS (Sequential Forward Selection), SBS (Sequential Backward Selection), SFFS (Sequential Floating Forward Selection), etc. are widely used as traditional feature selection techniques. It is also true that if there are many risks and many features, there is a limitation in that the performance for classification prediction is poor and it takes a lot of learning time. Therefore, in this study, we propose an improved feature selection algorithm to enhance the effectiveness of the existing campaign. The purpose of this study is to improve the existing SFFS sequential method in the process of searching for feature subsets that are the basis for improving machine learning model performance using statistical characteristics of the data to be processed in the campaign system. Through this, features that have a lot of influence on performance are first derived, features that have a negative effect are removed, and then the sequential method is applied to increase the efficiency for search performance and to apply an improved algorithm to enable generalized prediction. Through this, it was confirmed that the proposed model showed better search and prediction performance than the traditional greed algorithm. Compared with the original data set, greed algorithm, genetic algorithm (GA), and recursive feature elimination (RFE), the campaign success prediction was higher. In addition, when performing campaign success prediction, the improved feature selection algorithm was found to be helpful in analyzing and interpreting the prediction results by providing the importance of the derived features. This is important features such as age, customer rating, and sales, which were previously known statistically. Unlike the previous campaign planners, features such as the combined product name, average 3-month data consumption rate, and the last 3-month wireless data usage were unexpectedly selected as important features for the campaign response, which they rarely used to select campaign targets. It was confirmed that base attributes can also be very important features depending on the type of campaign. Through this, it is possible to analyze and understand the important characteristics of each campaign type.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.4A
• /
• pp.332-340
• /
• 2010

In this paper, we deal with multi-user diversity scheduling methods that transmit simultaneously signals from multiple users using superposition coding multiplexing. These methods can make various scheduling methods be obtained, according to strategies for user selection priority from the first user to the first-following users, strategies for per-user power allocation, and resulting combining strategies. For the first user selection, we consider three strategies such as 1) higher priority for a user with a better channel state, 2) following the proportional fair scheduling (PFS) priority, 3) higher priority for a user with a lower average serving rate. For selection of the first-following users, we consider the identical strategies for the first user selection. However, in the second strategy, we can decide user priorities according to the original PFS ordering, or only once an additional user for power allocation according to the PFS criterion by considering a residual power and inter-user interference. In the strategies for power allocation, we consider two strategies as follows. In the first strategy, it allocates a power to provide a permissible per-user maximum rate. In the second strategy, it allocates a power to provide a required per-user minimum rate, and then it reallocates the residual power to respective users with a rate greater than the required minimum and less than the permissible maximum. We consider three directions for scheduling such as maximizing the sum rate, maximizing the fairness, and maximizing the sum rate while maintaining the PFS fairness. We select the max CIR, max-min fair, and PF scheduling methods as their corresponding reference methods [1 and references therein], and then we choose candidate scheduling methods which performances are similar to or better than those of the corresponding reference methods in terms of the sum rate or the fairness while being better than their corresponding performances in terms of the alternative metric (fairness or sum rate). Through computer simulations, we evaluate the sum rate and Jain’s fairness index (JFI) performances of various scheduling methods according to the number of users.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_2
• /
• pp.1553-1563
• /
• 2023

The Geostationary Ocean Color Imager-II (GOCI-II) is a satellite designed for ocean color observation, covering the Northeast Asian region and the entire disk of the Earth. It commenced operations in 2020, succeeding its predecessor, GOCI, which had been active for the previous decade. In this study, we aimed to enhance the atmospheric correction algorithm, a critical step in producing satellite-based ocean color data, by performing cross-calibration on the GOCI-II near-infrared (NIR) band using the GOCI NIR band. To achieve this, we conducted a cross-calibration study on the top-of-atmosphere (TOA) radiance of the NIR band and derived a vicarious calibration gain for two NIR bands (745 and 865 nm). As a result of applying this gain, the offset of two sensors decreased and the ratio approached 1. It shows that consistency of two sensors was improved. Also, the Rayleigh-corrected reflectance at 745 nm and 865 nm increased by 5.62% and 9.52%, respectively. This alteration had implications for the ratio of Rayleigh-corrected reflectance at these wavelengths, potentially impacting the atmospheric correction results across all spectral bands, particularly during the aerosol reflectance correction process within the atmospheric correction algorithm. Due to the limited overlapping operational period of GOCI and GOCI-II satellites, we only used data from March 2021. Nevertheless, we anticipate further enhancements through ongoing cross-calibration research with other satellites in the future. Additionally, it is essential to apply the vicarious calibration gain derived for the NIR band in this study to perform vicarious calibration for the visible channels and assess its impact on the accuracy of the ocean color products.