• Journal of Information Processing Systems
• /
• v.11 no.3
• /
• pp.389-405
• /
• 2015

Most traditional database systems exploit a record-oriented model where the attributes of a record are placed contiguously in a hard disk to achieve high performance writes. However, for read-mostly data warehouse systems, the column-oriented database has become a proper model because of its superior read performance. Today, flash memory is largely recognized as the preferred storage media for high-speed database systems. In this paper, we introduce a column-oriented database model based on flash memory and then propose a new column-aware flash indexing scheme for the high-speed column-oriented data warehouse systems. Our index management scheme, which uses an enhanced $B^+$-Tree, achieves superior search performance by indexing an embedded segment and packing an unused space in internal and leaf nodes. Based on the performance results of two test databases, we concluded that the column-aware flash index management outperforms the traditional scheme in the respect of the mixed operation throughput and its response time.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2000.04a
• /
• pp.179-183
• /
• 2000

By spread of Internet, we become to use several of communication such as email, ICQ, VoIP etc. But as become convenient to use, user must be aware of variety of information regarding the media and partner user. The forecast is that this problem grows larger and larger as new media are brought in the expanding communications network. In this research, we suggest an agent MIA (Media Integration Agent) that manages the information instead of user. The MIA is an agent that obtains situation of user, and dynamically exchanged addresses and tendency of using medias in form of vCard. By use of this, user can choose better communication media. Accordingly seamless communication environment that user doesn't have to be aware of various information is formed.

• Proceedings of the Korea Society of Design Studies Conference
• /
• 2004.10a
• /
• pp.178-179
• /
• 2004

• Proceedings of the Korea Society of Design Studies Conference
• /
• 2005.10a
• /
• pp.116-117
• /
• 2005

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.14 no.1
• /
• pp.265-273
• /
• 2014

In this paper, context-aware technologies using a Kinect sensor for a digital signage which is increasingly growing as the 4th screen media is presented. Generalized context-aware functions for a digital signage were studied and a context-aware digital signage platform was developed. It can support a natural user interface for controlling a digital signage and actively provide contents adapted to its context. As a basic context, user's gesture and voice, the number of users, sound direction were considered. In the future, the advanced functions such as age, gender will be studied. The implemented platform could be a good reference model when developing a general context-aware digital signage.

• Journal of Broadcast Engineering
• /
• v.23 no.1
• /
• pp.26-35
• /
• 2018

The recent development of overlay network technology enables distributed sharing of various types of contents. Although overlay network has great advantages as a huge content repository, it is practically difficult to directly provide such Internet service as streaming of contents. On the other hand, the media server, which is specialized in content services, has excellent service capabilities, but it suffers from the huge contents that are constantly created and requires large expansion of severs and storages, and thus requires much effort for efficient management of the huge repository. Hence, the association of an overlay network of huge storage with a media server of high performance content service will show a great synergy effect. In this paper, a location-aware scheme of constructing overlay networks and associating it with media server is proposed, and then cache-based contents management and service policy are proposed for efficient content service. The performance is analysed for one of the content services, streaming service.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.130-135
• /
• 2006

본 논문에서는 상황 인지(Context Aware)를 이용한 제스처 인식 방법에 대하여 기술한다. 기존의 인식 방법들은 대부분 제스처의 개별적인 의미를 중심으로 제스처를 분류하는 방법을 사용한다. 그러나 이러한 방법들은 인식 알고리즘을 일반화하는데 있어서 다음과 같은 문제점들을 가지고 있다. 첫째, 인간의 모든 제스처를 제한된 특징으로 모호하지 않게 구별하기 어렵다. 둘째, 같은 제스처라 할지라도 상황에 따라 다른 의미를 내포할 수 있다. 이러한 문제점들을 해결하고자 본 논문에서는 확률 기반의 상황 인지 모델을 이용한 제스처 인식 방법을 제안한다. 이 방법은 제스처의 개별적인 의미를 인식하기 전에 대상의 상황을 추상적으로 분류함으로써 행위자의 의도를 정확히 파악할 수 있다. 본 방법은 시스템의 상태를 [NULL], [OBJECT], [POSTURE], [GLOBAL], [LOCAL]의 5 가지 상태로 정의한 뒤, 각 상태의 천이를 바탕으로 대상의 상황을 판단한다. 이러한 상황 정보에 따라 각 상태에 최적화된 인식 알고리즘을 적용함으로써 지능적인 제스처 인식을 수행할 수 있으며, 기존 방법들이 갖는 제스처 인식의 제약을 완화 시키는 효과가 있다. 따라서, 제안하는 제스처 인터페이스는 자연스러운 상호 작용이 필요한 지능형 정보 가전 혹은 지능형 로봇의 HCI 로 활용될 수 있을 것이다.

• Smart Media Journal
• /
• v.11 no.5
• /
• pp.82-90
• /
• 2022

Along with the growing popularity of 5G technology, providing flexible and personalized network services suitable for requirements of customers has also become a lucrative venture and business key for network service providers. Therefore, dynamic network provisioning is needed to help network service providers. Moreover, increasing user demand for network services meets specific requirements of users, including location, usage duration, and QoS. In this paper, a routing algorithm, which makes routing decisions using Reinforcement Learning (RL) based on the information about link stability, is proposed and called Link Stability aware Reinforcement Learning (LSRL) routing. To evaluate this algorithm, several mininet-based experiments with various network settings were conducted. As a result, it was observed that the proposed method accepts more requests through the evaluation than the past link annotated shorted path algorithm and it was demonstrated that the proposed approach is an appealing solution for dynamic network provisioning routing.

• Journal of KIISE:Information Networking
• /
• v.34 no.5
• /
• pp.379-388
• /
• 2007

With proliferation of computer and wireless network technology, it is common to access Internet through wireless network using mobile device. Ratio of using streaming media out of many applications through Internet is increasing not only in wired network but also in wireless network. Streaming media is much bigger than other contents and requires more network bandwidth and more computing resources. However mobile devices hate relatively poor computing resource and low network bandwidth. If media streaming service is provided for mobile devices without any consideration about network bandwidth and computing power, it is hard for the client to get high qualify service. Since mobile device is supported with very limited energy from the battery, media streaming should be adjusted to varying energy state of mobile device in realtime to ensure complete playback of streaming media. In this paper, we propose DFRC to provide high qualify service to mobile client through wireless network by controlling the number of frames transmitted to client based on computing resource and energy state of mobile device.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.709-713
• /
• 2009

Recently, a lot of research that applies data acquired from devices such as cameras and RFIDs to context aware services is being performed in the field on Life-Log and the sensor network. A variety of analytical techniques has been proposed to recognize various information from the raw data because video and audio data include a larger volume of information than other sensor data. However, manually watching a huge amount of media data again has been necessary to create supervised data for the update of a class or the addition of a new class because these techniques generally use supervised learning. Therefore, the problem was that applications were able to use only recognition function based on fixed supervised data in most cases. Then, we proposed a method of acquiring supervised data from a video sharing site where users give comments on any video scene because those sites are remarkably popular and, therefore, many comments are generated. In the first step of this method, words with a high utility value are extracted by filtering the comment about the video. Second, the set of feature data in the time series is calculated by applying functions, which extract various feature data, to media data. Finally, our learning system calculates the correlation coefficient by using the above-mentioned two kinds of data, and the correlation coefficient is stored in the DB of the system. Various other applications contain a recognition function that is used to generate collective intelligence based on Web comments, by applying this correlation coefficient to new media data. In addition, flexible recognition that adjusts to a new object becomes possible by regularly acquiring and learning both media data and comments from a video sharing site while reducing work by manual operation. As a result, recognition of not only the name of the seen object but also indirect information, e.g. the impression or the action toward the object, was enabled.