• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.5
• /
• pp.431-436
• /
• 2015

As a number of TV programs broadcast today, researches about TV program recommender system have been studied and many researchers have been studying recommender system to produce recommendation with high accuracy. Recommender system recommends TV program to user by using metadata like genre, plot or calculating users' preferences about TV programs. In this paper, we propose a new TV program Collaborative Filtering Recommender System that exploits viewing time pattern like viewing ratio, relation with finish time and recently viewing history to calculate preference for high-quality of recommendation. To verify usefulness of our research, we also compare our method which utilizes viewing time patterns and baseline which simply recommends TV program of user's most frequently watched channel. Through this experiments, we show that our method very effectively works and recommendation performance increases.

• The Journal of the Korea Contents Association
• /
• v.15 no.4
• /
• pp.153-167
• /
• 2015

This paper empirically analyzes specific characteristics of VOD users and their practical viewing patterns, using socio-demographic data of the current IPTV subscribers in combination with actual data of genre usage and payment. The findings revealed active viewing patterns of male, the unemployed, high earners and early adopters. In terms of preferences, households with large numbers of women prefer time shift contents, whereas households composed of more men or preschoolers prefer non-time shift contents. Likewise, the households that have more women or higher income had relatively a lot of experiences of purchasing time shift contents on one hand, but the households characterized by the larger numbers of family members or unemployed householder or young householder showed much willingness to pay premium contents on the other hand. Given the utilization of correct database, the findings offer useful information conducive to service promotion and marketing strategies to maximize VOD use in the practical dimension.

• Journal of Broadcast Engineering
• /
• v.14 no.5
• /
• pp.625-637
• /
• 2009

With advent of TV environment and increasing of variety of program contents, users are able to experience more various and complex environment for watching TV contents. According to the change of content watching environment, users have to make more efforts to choose his/her interested TV program contents or TV channels than before. Also, the users usually watch the TV program contents with their own regular way. So, in this paper, we suggests personalized TV program schedule recommendation system based on the analyzing users' TV watching history data. And we extract the users' watched program patterns using the sequential pattern mining method. Also, we proposed a new sequential pattern mining which is suitable for TV watching environment and verify our proposed method have better performance than existing sequential pattern mining method in our application area. In the future, we will consider a VoD characteristic for extending to IPTV program schedule recommendation system.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.105-110
• /
• 2009

Conventional TV viewing environment had provided limited numbers of channels and contents so that accessibility of contents was made user's manual change of TV channels and by manual selection of TV program contents. However, with advent of IPTV and various contents and channels available to users’ terminals, excessive numbers of TV contents become available to users’ terminals, thus leading to totally different TV viewing environments. In this TV environment, users are required to make much effort to choose their preferred TV channels or program contents, which becomes much cumbersome to the users. Therefore, in this paper, we will propose TV contents schedule recommendation by making reasoning on users’ TV viewing patterns from TV viewing history data using sequential pattern mining so that so that it increases accessibility of users to many TV program contents which may be or may not be aware of the users.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.135-136
• /
• 2018

최근 개인방송은 시청자가 언제 어디서나 원하는 장소, 시간에 시청할 수 있을 정도로 보편적이고 대중화 되어 있다. 일반적으로 시청자는 방송을 시청할 때 영상의 구간별로 진행되는 콘텐츠를 처음부터 시청하면서 확인하거나 댓글을 확인하기 전까지는 확인할 수 있는 방법이 없다. 이에 본 논문은 행동 패턴을 분석하여 해당 영상의 콘텐츠를 파악하여 시청자에게 방송진행자가 무엇을 하고 있는지를 제공할 수 있다. 따라서 시청자가 영상의 구간별 콘텐츠를 확인하여 영상에서 원하는 구간을 선택하여 시청이 가능하게 되면서 시청자에게 영상을 시청하는데 편리성을 줄 수 있는 기대효과를 가진다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.4
• /
• pp.537-542
• /
• 2021

Smart TVs provide a variety of services and information compared to existing TVs based on the Internet. In order to provide more personalized services or information, it is necessary to analyze users' viewing patterns and provide customized services or information based on them. The proposed system receives the user's TV viewing pattern, analyzes it, and recommends a TV program or movie as customized information to the user. For this, the system was constructed with a preprocessor and a deep learning model. The preprocessor refines the name of the TV program watched by the user, the date the TV program was watched, and the watched time. Then, the multi-attribute LSTM model trains the refined data and performs prediction.The proposed system is a system that provides customized information to users, and is believed to be a leading technology in digital convergence that combines existing IoT technology and deep learning technology.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.4
• /
• pp.1000-1006
• /
• 2014

This study focuses on the fact that a OTT service is vigorously used for smart phones, and suggests a design of method to trace the experiences of watching television contents. For this purpose, we developed logging functions and embedded them into existing OTT service application to record flow and pattern of watching context. This paper suggests a log file format which can accurately and precisely record watching actions of users per-second methodology rather than former per-minute methodology. Moreover, this study shows that the application can trace watching attitude according to occurring events by characteristics and playing modes of realtime broadcasting, VOD, advertisement contents. In addition, based on the result of the study, this paper discusses educational, operational meaning of the method such as methodological application in mobile ethnography field or survey for total screening rate.

• Journal of Advanced Navigation Technology
• /
• v.15 no.6
• /
• pp.1163-1172
• /
• 2011

This study attempted to analyze between TV and family. While almost research has focused on the TV influencing family, this study has tried to contemplate family type and family communication pattern influencing TV use. Especially, this research focused on adolescent TV use. The result and conclusion are as follow: There are different children's TV watching styles based on family communication pattern(conversation-orientation and conformity-orientation). The amount of TV watching is high in conformity-orientation family pattern, and the amount of TV watching is low in conversation-orientation family pattern. In addition, conformity-orientation family pattern more prefer entertainment program, while conversation-orientation family pattern prefer informational program.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2004.11a
• /
• pp.151-154
• /
• 2004

디지털 방송의 시작과 함께, 지상파, 위성, 케이블과 같은 다양한 매체를 통한 다채널 방송 시청 환경의 도래는 사용자에게 많은 방송 프로그램 시청 정보를 전달하게 되었다. 이와 더불어, 방송 단말에 전송된 다양한 방송 프로그램 정보를 탐색하고 선호 방송 프로그램을 선별하기 위해서는 사용자에게 많은 노력이 요구된다. 이러한 요구에 따라, 똔 논문에서는 다채널 방송 시청 환경 하에서 사용자의 방송 프로그램 시청 히스토리를 분석하고, 특정 시간에 따른 사용자의 방송 프로그램 시청 패턴을 추출하여 방송 프로그램 장르에 대한 사용자 선호도를 자동으로 계산하는 알고리즘을 제안하고, MPEG-7 MDS 구조에 따른 사용자 선호도 서술과 사용자의 선호도에 따라 방송 프로그램을 자동적으로 추천하는 TV 프로그램 추천 어플리케이션을 소개한다 본 실험을 위해 실제 연령대별, 성별, 시간대별로 사용자의 TV 시청 자료를 사용하였으며, 실험결과를 통해 본 논문에 제안된 베이시안 네트워크 기반 사용자 자동 학습 알고리즘이 효과적으로 사용자 선호도를 학습할 수 있음을 확인하였다.

• The Journal of the Korea Contents Association
• /
• v.10 no.1
• /
• pp.181-191
• /
• 2010

The number of subscribers of Broad & TV, Qook TV and My LG TV has been growing rapidly. It has been achieved at 1.7 million on August 2009. This study will focus on the viewing pattern of IP-VOD subscribers. For this research, we have interviewed 34 VOD users. According to the analyse, the VOD users prefer to watch free TV's programs, particularly variety programs, during 2 or 3 hours per view and 2 or 3 times a week. They use in general VOD service alone unlike the traditional TV viewers. Furthermore, There were single viewers who have used more than 8 hours continuously on weekend.