• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.11a
• /
• pp.174-176
• /
• 2017

매년 여행을 즐기는 여행객들의 수가 꾸준히 증가하고 있으며, 이러한 추세는 해외여행 뿐 아니라 국내 여행에서도 나타나고 있다. 국내 여행을 즐기는 여행객 수의 증가는 매우 다양하고 복합적인 요인들에 의해 이루어지고 있는 것이 사실이나, 국내 여행객들의 절대 다수는 해외여행과는 달리 패키지 형태 보다는 자유여행 형태의 여행을 선호하고 있다. 이는 해외 여행지 대비 국내 여행지가 여행객들이 취득 및 분석할 수 있는 정보의 접근성이 훨씬 높고 정보의 양 역시 풍부하다는 것에서 기인한다고 할 수 있다. 그러나 이러한 정보 접근의 용이성 및 정보량의 풍요성은 오히려 자유여행을 즐기고자 하는 여행객들이 여행코스 및 숙소를 정하는데 많은 시간을 투자하게 되는 요인으로 작용하고 있다. 때문에 이러한 단점을 해결하고자 본 논문에서 제안하는 애플리케이션은 국내 여행객들이 편리하고 손쉽게 국내 여행을 즐길 수 있도록 여행코스 및 숙소를 지정할 수 있는 기능을 제공한다. 이를 위해, 제안하는 애플리케이션에서는 국내 여행과 관련된 다양한 정보들을 각종 포털 사이트와 SNS에서 수집하고, 이를 기반으로 사용자 선호 정보와의 매칭을 통해 맞춤형 여행 코스 제안 및 숙소 예약 기능을 제공한다. 이를 통해, 국내 여행을 즐기는 여행객들에게 편리함을 제공하고, 국내 여행객 수의 증가를 기대할 수 있다.

• The KIPS Transactions:PartD
• /
• v.14D no.3 s.113
• /
• pp.339-346
• /
• 2007

The objective of this study is to satisfy requirements for a variety of terminals to play wired web page contents in ubiquitous environment constantly connected to network. In other words, this study intended to automatically transcode wired web page into mobile web page in order to receive service by using mobile base to carry contents in Internet web page. To achieve this objective, we suggest the method that is to directly enter URL of web page in mobile device and to check contents of the current web page. For this, web page is converted into an image and configured into a mobile web page suitable for personal terminals. Users can obtain the effect of having web services provided by using computer with interfaces to expand, reduce and move the web page as desired. This is a caret unit play method, with which contents of web page are transcoded and played to suit each user According to the method proposed in this study, contents of wired web page can be played by using a mobile device. This study confirms that a single content can be serviced to suit users of various terminals. Through this, it will be able to reuse numerous wired web contents as mobile web contents.

• The KIPS Transactions:PartD
• /
• v.10D no.4
• /
• pp.577-590
• /
• 2003

The shape-based retrieval is defined as the operation that searches for the (sub) sequences whose shapes are similar to that of a query sequence regardless of their actual element values. In this paper, we propose a similarity model suitable for shape-based retrieval and present an indexing method for supporting the similarity model. The proposed similarity model enables to retrieve similar shapes accurately by providing the combination of various shape-preserving transformations such as normalization, moving average, and time warping. Our indexing method stores every distinct subsequence concisely into the disk-based suffix tree for efficient and adaptive query processing. We allow the user to dynamically choose a similarity model suitable for a given application. More specifically, we allow the user to determine the parameter p of the distance function $L_p$ when submitting a query. The result of extensive experiments revealed that our approach not only successfully finds the subsequences whose shapes are similar to a query shape but also significantly outperforms the sequence search.

• Journal of the Korea Computer Graphics Society
• /
• v.26 no.3
• /
• pp.87-97
• /
• 2020

In this paper, we propose a novel neural relighting that infers a relighted rendering image based on the user-guided specular highlight map. The proposed network utilizes a pre-trained neural renderer as a backbone network learned from the rendered image of a 3D scene with various lighting conditions. We jointly optimize a 3D light position and its associated relighted image by back-propagation, so that the difference between the base image and the relighted image is similar to the user-guided specular highlight map. The proposed method has the advantage of being able to explicitly infer the 3D lighting position, while providing the artists' preferred 2D screen-space interface. The performance of the proposed network was measured under the conditions that can establish ground truths, and the average error rate of light position estimations is 0.11, with the normalized 3D scene size.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.6
• /
• pp.643-656
• /
• 2019

The development of hot social networking services, including the Internet, has transformed rental car services into IT-based car-sharing services. The car-sharing service is a service that allows people to rent cars online without having to face-to-face. It has become the preffered service among young people who are accustomed to drinking and eating alone. Users can use their smartphones to book their types of cars and hours of rent, and then go to a nearby designated parking lot to use their reserved cars. You can open a designated car door with a smart phone and drive. It is a very convenient service to pay for the distance you drive. However, the car-sharing service already in use in business has the following problems: underage who do not have a driver's license may drive a car borrowed by an acquaintance, the status of a license registered at the time of join membership with a car-sharing company may change to a suspension or cancellation of a license while renting and driving, or even a drunk person may rent a car and drive. In this paper, the method to solve these problems has been studied and proposed. The proposed method is to reduce the cost of investment by a car-sharing service provider and to minimize user inconvenience. And, it was compared and analyzed with the existing method. For the method to be used efficiently, the active operation of the car-sharing company and the government's policies will have to be supported.