• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.53-65
• /
• 2019

Object tracking is one of important steps to achieve video-based surveillance systems. Object tracking is considered as an essential task similar to object detection and recognition. In order to perform object tracking, various machine learning methods (e.g., least-squares, perceptron and support vector machine) can be applied for different designs of tracking systems. In general, generative methods (e.g., principal component analysis) were utilized due to its simplicity and effectiveness. However, the generative methods were only focused on modeling the target object. Due to this limitation, discriminative methods (e.g., binary classification) were adopted to distinguish the target object and the background. Among the machine learning methods for binary classification, total error rate minimization can be used as one of successful machine learning methods for binary classification. The total error rate minimization can achieve a global minimum due to a quadratic approximation to a step function while other methods (e.g., support vector machine) seek local minima using nonlinear functions (e.g., hinge loss function). Due to this quadratic approximation, the total error rate minimization could obtain appropriate properties in solving optimization problems for binary classification. However, this total error rate minimization was based on a batch mode setting. The batch mode setting can be limited to several applications under offline learning. Due to limited computing resources, offline learning could not handle large scale data sets. Compared to offline learning, online learning can update its solution without storing all training samples in learning process. Due to increment of large scale data sets, online learning becomes one of essential properties for various applications. Since object tracking needs to handle data samples in real time, online learning based total error rate minimization methods are necessary to efficiently address object tracking problems. Due to the need of the online learning, an online learning based total error rate minimization method was developed. However, an approximately reweighted technique was developed. Although the approximation technique is utilized, this online version of the total error rate minimization could achieve good performances in biometric applications. However, this method is assumed that the total error rate minimization can be asymptotically achieved when only the number of training samples is infinite. Although there is the assumption to achieve the total error rate minimization, the approximation issue can continuously accumulate learning errors according to increment of training samples. Due to this reason, the approximated online learning solution can then lead a wrong solution. The wrong solution can make significant errors when it is applied to surveillance systems. In this paper, we propose an exactly reweighted technique to recursively update the solution of the total error rate minimization in online learning manner. Compared to the approximately reweighted online total error rate minimization, an exactly reweighted online total error rate minimization is achieved. The proposed exact online learning method based on the total error rate minimization is then applied to object tracking problems. In our object tracking system, particle filtering is adopted. In particle filtering, our observation model is consisted of both generative and discriminative methods to leverage the advantages between generative and discriminative properties. In our experiments, our proposed object tracking system achieves promising performances on 8 public video sequences over competing object tracking systems. The paired t-test is also reported to evaluate its quality of the results. Our proposed online learning method can be extended under the deep learning architecture which can cover the shallow and deep networks. Moreover, online learning methods, that need the exact reweighting process, can use our proposed reweighting technique. In addition to object tracking, the proposed online learning method can be easily applied to object detection and recognition. Therefore, our proposed methods can contribute to online learning community and object tracking, detection and recognition communities.

• Asia pacific journal of information systems
• /
• v.17 no.4
• /
• pp.165-186
• /
• 2007

With the advent of a wide variety of Internet shopping malls, consumers can choose a best appealing shopping mall from among the Bricks-and-Clicks and Pure-Player malls. Pure-Players launched their operation grandiosely with the early stage of Internet use in 1995. However, after the burst of Dot-com company bubbles in 1997, Pure-Players introduce various types of business models to meet potential needs of consumers. While Pure-Players suffer skeptical views from market analysts as well as consumers, traditional offline companies learned important lessons from Dot-com companies collapse phenomena, and expanded their business channels into online in the name of Bricks-and-Clicks. Nowadays, Bricks-and-Clicks successfully establish in the market as one of reliable business partners among consumers. Therefore, it is no surprise that recent competitions between Bricks-and Clicks and Pure-Players become fiercer than ever to attract potential customers to their websites. In this situation, consumers can choose a shopping mall to their best satisfaction. Consumers can enjoy both offline and online options for shopping because Bricks-and Clicks provide both offline and online channels to consumers, which is compared with Pure-Players offering only online channel. Offline channel is unique in providing consumers with chances to touch and feel target products and services. Meanwhile, online channel is considered very viable and convenient shopping options for consumers. In this respect, it is easily assumed that consumers will show different online shopping behavior when they have to choose either Bricks-and-Clicks mall or Pure-Player mall for the sake of shopping. Remaining research issue in this case is how much consumers' schema would influence online shopping behavior between Bricks-and-Clicks and Pure-Players. Basically, schema is a framework for synthetic information recognition that individual consumers have and is very characteristic in that it focuses not on fragmentary facts but on the combination of various causes affecting results. Consumers' schema is closely represented by trust, structural assurance, and perceived relative advantage towards a specific type of shopping mall. In literature, there exist a lot of studies comparing Bricks-and-Clicks and Pure-Players. However, there is no study to pursue the analysis of consumer behaviors comparing Bricks-and Clicks and Pure-Players from the schema perspective. Therefore, this study aims to investigate this research gap. Empirical analysis is adopted by garnering valid questionnaires from 514 Internet shopping mall users. 237 were mainly using Bricks-and-Clicks for shopping, while 277 were found to visit Pure-Players for shopping. PLS was applied to analyze the survey data to verify the proposed research hypotheses. Findings from the empirical test results are as follows. First, consumers perceive more trust and relative advantage in Pure-Players, comparing with Bricks-and-Clicks. This result is against widely-accepted perception that Bricks-and-Clicks would be perceived by consumers as more trustworthy and relatively advantageous because they have offline reputation and stores. Therefore, it becomes more obvious that Internet is becoming daily necessaries, and consumers increasingly feel very comfortable in using the Internet for their own personal purposes. Second, consumers have firm faith in transaction safety, regardless Bricks-and-Clicks and Pure-Players. This seems due to the fact that most of shopping malls showing dubious transaction safety have no place in the market. In a nutshell, empirical results tell us that Pure-Players will grow very much in the future, to the extent that consumers perceive no difference in comparison with Bricks-and-Clicks. Besides, consumers' schema accumulated through trust and perceived relative advantage plays crucial role in determining consumer behavior.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.8
• /
• pp.149-156
• /
• 2013

TAmong various forms of online shopping with the development of online shopping and the rapid growth of IT technology, the intense competition also more deepen. The strategies for differentiated marketing and features for competitiveness of businesses in the online and offline shopping areas are required that our marketing target be settled the flower marketing by analyzing. In case of the flower market, the movement of the market from offline market to online market has been occurred and processed concurrently. Market entry and competitiveness of small companies is becoming very difficult because larger companies have been formed national market networks. Therefore, in this research, we suggest ways to enhance flower delivery service to the market from the top ranked site analysis, market situation analysis and service strategy plan for the activation of the online flower market. Also, we describe how to realize the above mentioned strategies through a site building. We can expect to strengthen flower market activation and competitiveness through the flower delivery service enhancement.

• Korean Journal of Organic Agriculture
• /
• v.30 no.2
• /
• pp.191-206
• /
• 2022

Rice has prevailed as a staple food in South Korea. Rice consumption has sharply shrunk nowadays. In the wake of COVID-19 Shock and the "untact" era, the instant rice market has significantly grown, and there is a possibility of solving the shrinking consumption of domestic rice. This study contributes to the development of rice consumption promotion strategies by examining the factors of purchasing instant rice based on the Hedonic Price Methodology. In particular, given the increase in online market, this study compares attribute values of product characteristics between online and offline markets. The empirical results show that calorie, brand, organic and functional products have positive effects on instant rice prices. The rate of carbohydrate, the PB and bundle attributes have negative effects on instant rice prices. The results also show that the magnitude of brand, bundle and PB attribute values are bigger in offline market while that of the number of multigrain attribute value is relatively bigger in online market. The organic attribute value is important regardless of marketing channels.

• The Journal of Society for e-Business Studies
• /
• v.26 no.4
• /
• pp.81-93
• /
• 2021

Through processors, wholesale markets, intermediate sellers, and retailers, agricultural products have been distributed in a multi-level customary manner for a long time as they are easy to deteriorate and no not have a standardized system of size and quality. However, with the advancement of Internet networks and logistic services during the 2000s that facilitated the development of offline markets, and the rise of the non-contact purchase preference in direct response to COVID-19, previous offline consumers flowed into the online market to purchase agricultural goods. In other words, the volume of online agricultural transactions exploded since the pandemic. Against this social backdrop, this study focused on the difference in distribution costs as a result of converting from conventional offline distribution channels to online channels, and analyzed the reduced distribution costs through a case study of garlic sales on the online platform "H" shopping mall. The analysis found that considerable economic effects occurred, some of the effects being an approximate 39% decrease in distribution cost when comparing direct online transactions of the online shopping mall with other more traditional means, a reduced distribution cost rate of approximately 28%p, and increased profit for farmers.

• The Journal of Economics, Marketing and Management
• /
• v.11 no.1
• /
• pp.21-30
• /
• 2023

Purpose: As one of the ongoing studies on the distribution industry, the purpose of this study is to identify the research trends on online shopping so far to propose not only the development of online shopping companies but also the possibility of coexistence between online and offline retailers and the development of the distribution industry. Research design, data and methodology: In this study, the English abstracts of 645 papers on online shopping registered in scienceON were obtained. For the analysis through BERTopic and LDA using Python 3.7 and identifying which topics were interesting to researchers. Results: As a result of word frequency analysis and co-occurrence analysis, it was found that studies related to online shopping were frequently conducted on factors such as products, services, and shopping malls. As a result of BERTopic, five topics such as 'service quality' and 'sales strategy' were derived, and as a result of LDA, three topics including 'purchase experience' were derived. It was confirmed that 'Customer Recommendation' and 'Fashion Mall' showed relatively high interest, and 'Sales Strategy' showed relatively low interest. Conclusions: It was suggested that more diverse studies related to the online shopping mall platform, sales content, and usage influencing factors are needed to develop the online shopping industry.

• Journal of Korea Port Economic Association
• /
• v.35 no.1
• /
• pp.39-56
• /
• 2019

This study addresses the management performance of the retail industry in Korea. The Non-store e-commerce business, consisting of TV home shopping and online shopping, grew very fast during the 2000s. For roughly six years in the 2010s, the diversification of online shopping business models, intensifying competition, and the proliferation of competition accelerated the entry of offline shopping retailers. During the analysis period from 2001 to 2016, the performance of the Non-­store e-commerce business was statistically better than that of the offline shopping retail business, department stores, general retailers, and small­-sized retailers. As expected, a significant difference between performance and some of the indicators of growth and activity could be seen because of the capital and operating structure of the retail industry. In particular, the Non-store e-commerce businesses have diverse sales media and heterogeneous business models. In this study, we have presented the significance of the differences in the indicators of operating profit and inventory turnover, as well as the discussion required for managing in the future.

• Information Systems Review
• /
• v.18 no.2
• /
• pp.199-223
• /
• 2016

As social networking services (SNS) become widely used tools for maintaining social relationships, people use SNS to express themselves online. Users are free to form multiple characters in SNS because of online anonymity. This phenomenon causes SNS users to easily demonstrate multiple personas that are different from their identities in the real world. Therefore, this study focuses on online multi-personas that establish multiple fake identities in the SNS environment. The main objective of this study is to investigate factors that affect online multi-personas. Fake online identities can have various negative consequences such as cyber bullying, cyber vandalism, or antisocial behavior. Since the boundary between the online and offline worlds is fading fast, these negative aspects of online behavior may influence offline behaviors as well. This study focuses on teenagers who often create multi-personas online. According to previous studies, personal identities are usually established during a person's youth. Based on data on 664 teenage users, this study identifies four emotional factors, namely, closeness with others, relative deprivation, peer pressure and social norms. According to data analysis results, three factors (except closeness with others) have positive correlations with users' multi-personas. This study contributes to the literature by identifying the factors that cause young people to form online multi-personas, an issue that has not been fully discussed in previous studies. From a practical perspective, this study provides a basis for a safe online environment by explaining the reasons for creating fake SNS identities.

• Management Science and Financial Engineering
• /
• v.13 no.1
• /
• pp.55-72
• /
• 2007

We consider the product differentiation model of online channel competition and examine the strategies of hybrid firms in terms of efficiency. After measuring the social cost of online business strategies, we show that (i) online channel of hybrid firm under blockaded entry may increase the social cost if the firms' delivery cost is sufficiently smaller than the consumer's transportation cost, and (ii) online competition under free entry may increase the social cost if the firms' delivery cost is sufficiently larger than the consumer's transportation cost. Finally, we discuss the strategic incentive of hybrid firms to reduce delivery cost and investigate the effect of the Internet maturity on the social cost.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.12
• /
• pp.2226-2232
• /
• 2007

This paper presents online neural control approach for nonlinear mechanical systems with random friction nature. We construct neural auxiliary control to compensate a control error in online for overcoming friction effect which reduces control performance in real-time implementation. Friction dynamics is estimated by using online least square(LS) method, which is utilized for online learning of the neural network. We accomplish computer simulation for evaluating the proposed control approach comparing offline control method to demonstrate its superiority.