Jung K.-H.,SK Planet |
Lee J.,Seoul National University
Information Sciences | Year: 2013
As the volume of database grows, retrieval and ordering of information according to relevance has become an important and challenging task. Ranking problem has recently been considered and formulated as a machine learning problem. Among the various learning-to-rank methods, the ranking support vector machines (SVMs) have been widely applied in various applications because of its state-of-the-art performance. In this paper, we propose a novel ranking method based on a probabilistic generative model approach. The proposed method utilizes multi-support vector domain description (multi-SVDD) and constructs pseudo-conditional probabilities for data pairs, thus enabling the construction of an efficient posterior probability function of relevance judgment of data pairs. Results of experiments on both synthetic and real large-scale datasets show that the proposed method can efficiently learn ranking functions better than ranking SVMs. © 2013 Elsevier Inc. All rights reserved.
Moon H.,Pohang University of Science and Technology |
Han S.H.,Pohang University of Science and Technology |
Chun J.,Carnegie Mellon University |
Hong S.W.,SK Planet
Human Factors and Ergonomics In Manufacturing | Year: 2016
A customer journey map (CJM) is a widely used tool to represent user experience with a service. Although numerous companies have used this tool to improve existing services or to develop new services, the maps are neither consistent nor mutually compatible because no clear design process for a CJM has been presented. This study aimed to develop a design process and rule sets for a CJM based on a human factors approach. The 10-step process and the rule sets were built on case studies of 25 categories of mobile services. Large-scale case studies were conducted with mobile service providers and combined with the result of a user diary method that collected users' daily activities and the difficulties that the user perceived when performing a task. We suggested various methods for using the CJM to generate new service opportunities. The proposed design process and the way for opportunity discovery can help service designers to develop unified CJMs and to identify innovative service ideas. © 2016 Wiley Periodicals, Inc.
Kim J.,KIPFA |
Kim S.,Naver Corporation |
Beck S.,Daum Kakao |
Jung K.,SK Planet
Conference on Human Factors in Computing Systems - Proceedings | Year: 2015
Why foreign companies are not successful in portal business in Korean market? We believe that there are at least three reasons for the phenomenon. First, domestic firms have their own methods and skills to collect users' needs and wants. Second, the domestic firms have efficient internal process of fast implementation. Finally, the role and responsibility of domestic UX teams are rather peculiar to the contexts of domestic internet market. This panel will be presented by the UX professionals from the three major internet portals in Korea. The panel will be presided by the president of Korean Internet Professional Association. Korea Internet Professionals Association, an affiliate of 'Ministry of Science, ICT and Future Planning', which researches Internet technologies, services, planning, design, development, design, promotion, advertising, marketing, analysis, contracting, purchasing, procurement, testing, supervision, evaluation, inspection, management, consulting and education using wired and wireless Internet is a non-profit organization which consists of Internet companies and Internet experts. Copyright is held by the author/owner(s).
Choi J.-H.,Korea University |
Yun M.-J.,Real D Square |
Kim A.-R.,SK Planet |
Kim J.-O.,Korea University
3DTV-Conference | Year: 2012
The 2D/3D converting industry and its accompanying researches are rapidly growing. However, some of the converted 3D contents are low-quality and may be not viable for consumers. Quality Measure (QM) factors of the 2D/3D converted contents are depth perception, visual comfort and erroneousness. In this paper, we particularly focus on evaluating visual comfort among stereo quality measure factors. For this evaluation, we identify 8 factors to capture spatiotemporal characteristics of disparity. The final prediction of visual comfort is determined by linearly combining the 8 factors. According to experimental results, our objective prediction is highly correlated with subjective evaluation. © 2012 IEEE.
Kim I.-H.,SK Planet |
Shin S.Y.,Kumoh National Institute of Technology |
Song Y.-H.,SK Planet |
Chung K.-D.,National School in Computer Science |
Jeong G.-M.,Kookmin University
Journal of Electrical Engineering and Technology | Year: 2013
In this paper, we propose a capacity modeling method for Bluetooth based LBSs (Location Based Services) between Bluetooth APs(Access Points) and mobile phones. In order to do this, we consider the entrance and exit of the users to and from the AP. Using parameters such as the connection time, data transmission time and number of users in the AP zone, we analyze the capacity of a Bluetooth AP using an appropriate queuing model. The required number of APs in a certain zone is estimated from the number of users in the zone using the proposed queuing model. By performing simulations and experiments, the validity and applicability of the proposed method are verified.