Sanggil Kang

Target advertisement service using TV viewers’ profile inference

Due to the limitation of broadcasting service, in general, TV programs with commercial advertisements are scheduled to be broadcasted by demographics. The uniformly provided commercial can not draw many TV viewers’ interest, which is not correspondent to the goal of the commercial. In order to solve the problem, a novel target advertisement technique is proposed in this paper. The target advertisement is a personalized advertisement according to TV viewers’ profile such as their age, gender, occupation, etc. However, viewers are usually reluctant to inform their profile to the TV program provider or the advertisement company because their information can be used on some bad purpose by unknown people. Our target advertisement technique estimates a viewer’s profile using Normalized Distance Sum and Inner product method. In the experiment, our method is evaluated for estimating the TV viewers’ profile using TV usage history provided by AC Neilson Korea.

Statistical inference method of user preference on broadcasting content

This paper proposes a novel approach for estimating the statistical multimedia user preference by providing weights to multimedia contents with respective to their consumed time. The optimal weights can be obtained by training the statistical system in the sense that the mutual information between old preference and current preference is maximized. The weighting scheme can be done by partitioning a users consumption history data into smaller sets in a time axis. With developing a mathematical derivation of our learning method, experiments were implemented for predicting the TV genre preference using 2,000 TV viewers watching history and showed that the performance of our method is better than that of the typical method.

A novel personalized paper search system

In this paper we propose a novel personalized paper search system using the relevance among user’s queried keywords and user’s behaviors on a searched paper list. The proposed system builds user’s individual relevance network from analyzing the appearance frequencies of keywords in the searched papers. The relevance network is personalized by providing weights to the appearance frequencies of keywords according to users’ behaviors on the searched list, such as “downloading,” “opening,” and “no-action.” In the experimental section, we demonstrate our method using 100 faculties’ search information in the University of Suwon.

Personalized digital e-library service using users’ profile information

We propose a personalized digital E-library system using a collaborative filtering technique, which provides a personalized search list according to users’ preference. The proposed system analyzes the registered users’ actions such as “clicking” and “borrowing” items. According to the different actions, we provide a weight for calculating the users’ preference of each item. However, the list is uniformly provided to the individual users when they search with same keywords. In order to avoid the problem, we customize the order of items in the list according to whether there is any mismatching of profiles among registered users and target users or not.

Agent-based flexible videoconference system with automatic QoS parameter tuning

In this paper, we propose a new agent-based flexible videoconference system (AVCS) by modifying videoconference manger (VCM) agent in a conventional flexible videoconference system (FVCS). The proposed AVCS can more flexibly cope with changes in working conditions during videoconferencing than the conventional FVCS. It is because an automatic parameter tuning algorithm is imbedded to VCM dynamically adapt QoS (Quality of Service) parameters in the sense that the current working condition can meet with the desired working condition of the user, which can change in time during videoconferencing. In the experimental section, we design a new structure of the VCM with the automatic parameter tuning module, imbed to the prototype of FVCS and implement the new AVCS. Also, it is shown that the proposed AVCS outperforms the existing FVCS in the experiment.

Image Identification System Using MPEG-7 Descriptors

In this paper, we propose an image identification system employing MPEG-7 as feature and using neural network as the main methodology. The proposed system properly identifies whether a given image belongs to particular sports images or not. The simulation results show that the proposed system successfully identifies images with the rate of over 85%.

Dynamical e-commerce system for shopping mall site through mobile devices

We introduce a novel personalized E-commerce system through mobile devices. By providing mobile clients preferred service category or items in a shopping mall website, the problem of the limitation of resource of mobile devices can be solved. In this paper, the preferred service items are inferred by analyzing customers statistical preference transactions and consumption behaviors in the website. In computing the statistical preference transactions, we consider the ratio of the length of each service page and customers staying time on it. Also, our system dynamically provides the personalized E-commerce service according to the three different cases such as the beginning stage, the positive response, and the negative response. In the experimental section, we demonstrate our personalized E-commerce service system and show how much the resource of mobile devices can be saved.

Dynamically personalized web service system to mobile devices

We introduce a novel personalized web service system through mobile devices. By providing only users preferred web pages or smaller readable sections, service elements, the problem of the limitation of resource of mobile devices can be solved. In this paper, the preferred service elements are obtained from the statistical preference transactions among web pages for each web site. In computing the preference, we consider the ratio of the length of each web page and users staying time on it. Also, our system dynamically provides the personalized web service according to the different three cases such as the beginning stage, the positive feedback, and the negative feedback. In the experimental section, we demonstrate our personalized web service system and show how much the resource of mobile devices can be saved.

Multi-module image classification system

In this paper, we propose an image classification system employing multiple modules. The proposed system hierarchically categorizes given sports images into one of the predefined sports classes, eight in this experiment. The image first categorized into one of the two classes in the global module. The corresponding local module is selected accordingly, and then used in the local classification step. By employing multiple modules, the system can specialize each local module properly for the given class feature. The simulation results show that the proposed system successfully classifies images with the correct rate of over 70%.

A Multimedia Contents Recommendation System using Preference Transition Probability

Recently Digital multimedia broadcasting (DMB) has been available as a commercial service. The users sometimes have difficulty in finding their preferred multimedia contents and need to spend a lot of searching time finding them. They are even very likely to miss their preferred contents while searching for them. In order to solve the problem, we need a method for recommendation users preferred only minimum information. We propose an algorithm and a system for recommending users` preferred contents using preference transition probability from user`s usage history. The system includes four agents: a client manager agent, a monitoring agent, a learning agent, and a recommendation agent. The client manager agent interacts and coordinates with the other modules, the monitoring agent gathers usage data for analyzing the user`s preference of the contents, the learning agent cleans the gathered usage data and modeling with state transition matrix over time, and the recommendation agent recommends the user`s preferred contents by analyzing the cleaned usage data. In the recommendation agent, we developed the recommendation algorithm using a user`s preference transition probability for the contents. The prototype of the proposed system is designed and implemented on the WIPI(Wireless Internet Platform for Interoperability). The experimental results show that the recommendation algorithm using a user`s preference transition probability can provide better performances than a conventional method.