Hyeong-Joon Kwon

Personalized smart TV program recommender based on collaborative filtering and a novel similarity method

The viewing set-based method has difficulties ensuring that a user will enjoy recommended programs, and the model-based collaborative filtering method contains system-side real-time recommendation problems because most recent ratings cannot be applied in the recommendations and it has increased calculating costs due to the training process. In this paper, we propose a personalized program recommender for smart TVs using memory-based collaborative filtering with a novel similarity method that is robust to cold-start conditions and faster than the often-used, existing similarity method. The proposed method can improve the recommendation performance of electronic program guides and recommender applications for smart TVs. We determined the prediction accuracy of the ratings under various conditions in order to evaluate the proposed method. As a result, we confirmed that the proposed method is effective for cold-start conditions.

Location awareness-based intelligent multi-agent technology

In this paper, we propose an advanced location awareness-based intelligent multi-agent technology that allows multiple users to share various user-centric mobile multimedia contents. This paper mainly focuses on (1) mobile station-based mixed-web map module via mobile mash-up technology, (2) a new location-based mobile multimedia technology using ubiquitous sensor Net.-based five senses content, and (3) location awareness-based intelligent multi-agent technology that includes a location-based integrated retrieval agent, a mobile social network (MSN)-based multi-user detection agent and user-centric automatic mobile multimedia recommender agent. This paper aims at validating and inspecting the usability, suitability, and applicability of the suggested technology via various performance evaluation experiments.

Improving Prediction Accuracy Using Entropy Weighting in Collaborative Filtering

In this paper, we evaluate performance of existing similarity measurement metric and propose a novel method using user’s preferences information entropy to reduce MAE in memory-based collaborative recommender systems. The proposed method applies a similarity of individual inclination to traditional similarity measurement methods. We experiment on various similarity metrics under different conditions,which include an amount of data and significance weighting from n/10 to n/60, to verify the proposed method. As a result, we confirm the proposed method is robust and efficient from the viewpoint of a sparse data set, applying existing various similarity measurement methods and Significance Weighting.

Mobile Web 2.0-Oriented Five Senses Multimedia Technology with LBS-Based Intelligent Agent

Ubiquitous-oriented realistic next generation mobile multimedia technology requires new approaches that sufficiently reflect humans sensory information and mobile Web 2.0-oriented collective intelligence and social networking concepts. Hence, we suggest and implement enhanced user location-based five senses multimedia technology that realizes a collective intelligence and mobile social networking between multi-mobile users. This includes 1) mobile station-based mixed-web map module via mobile mash-up, 2) authoring module of location-based five senses multimedia contents using ubiquitous-oriented sensor network and WiBro(Mobile Wi-Max), and 3) LBS-oriented intelligent agent module that includes ontology-based five senses multimedia retrieval module, social network-based user detection interface and user-centric automatic five senses multimedia recommender interface.

Personalized electronic program guide for IPTV based on collaborative filtering with novel similarity method

We propose a personalized electronic program guide (EPG) for IPTV. It uses memory-based collaborative filtering with a fast and accurate similarity method. The users explicit interests-based proposed method predicts users ratings on unexperienced content not previously rated by the user, and then arranges the content in order of highest ratings and classifies them according to their attributes. We experimented on the prediction accuracy of the ratings in order to evaluate the proposed method. As a result, we confirmed that the proposed method is effective for high-speed rating prediction and has improved accuracy.

Design and implementation of enhanced real time news service using RSS and VoiceXML

In ubiquitous computing, most people need to track various sources of news using various devices, but it becomes difficult once there are more than a handful of sources. This is the reason for this is that users have to navigate to each page, load it, remember how its formatted, and find where they last left off in the list. To solve these problems, many service providers provide RDF Site Summary documents. In this paper, we propose a newly designed news service using RSS and VoiceXML. RSS is an XML format that supports the syndication of news stories and similar content. There are several different formats for XML syndication that are referred to as RSS. Since RSS is in XML format, turning it into VoiceXML is easy and the synergy benefits of binding RSS and VXML are great. VoiceXML is a non-proprietary, web-based markup language for creating vocal dialogues between humans and computers. In this paper, we first focus on binding RSS and VXML, and RSS feed parsing. As a result of this research, we implement enhanced real time news service. People can use our service with their wired and wireless phone at any time, at any place. Also, we validate usability by comparing a typical RSS service scenario and typical VoiceXML service scenario, and calculating users satisfaction.

Multiple Odor Recognition and Source Direction Estimation with an Electronic Nose System

We propose an electronic nose system that can perform real time direction estimation of an odor source and multiple odors recognition based on a stereo sensor array for extensive use in mobile environments. The proposed system consists of the following: (1) a method to obtain odor signals using a twin-sensor array, which consists of 16-channel metal oxide semiconductor sensors; (2) a method to estimate the direction of an odor source by analyzing the signal amplitude of each channel in the stereo sensor array; and (3) a method to recognize two odors simultaneously using a hierarchical elimination method. We determine the accuracy of the direction estimation of odor sources and the odor recognition rate in order to verify the performance of the multiple odors recognition method. As a result, we confirm the high estimation performance of the model for the front three-way directions, with a recognition rate of approximately two odors simultaneously.

Novel Neighbor Selection Method to Improve Data Sparsity Problem in Collaborative Filtering

Memory-based collaborative filtering selects the top-k neighbors with high rank similarity in order to predict a rating for an item that the target user has not yet experienced. The most common traditional neighbor selection method for memory-based collaborative filtering is priority similarity. In this paper, we analyze various problems with the traditional neighbor selection method and propose a novel method to improve upon them. The proposed method minimizes the similarity evaluation errors with the existing neighbor selection method by considering the number of common items between two objects. The method is effective for the practical application of collaborative filtering. For validation, we analyze and compare experimental results between an existing method and the proposed method. We were able to confirm that the proposed method can improve the prediction accuracy of memory-based collaborative filtering by neighbor selection that prioritizes the number of common items.

Personalized real-time location-tagged contents recommender system based on mobile social networks

This paper proposes a real-time location-tagged contents recommender system which is based on mobile social network. The system locates a user via global positioning system, and then applies distance and preference filtering methods. We confirmed that the system is highly effective and applicable to convergence by a location data and content recommender through an implementation and preference prediction experiments.

Personalized Mobile Social Network System Using Collaborative Filtering

In this paper, we propose a location-based mobile social network system that integrates collaborative content recommendation The proposed system is an effective fusion of a traditional social network system, a location-based system, and a provider of intelligent recommendations. The prototype and service scenario in this study show possibilities for technical advancement and future extension of mobile social networking services. To verify the proposed system, we completed experiments to determine the recognition rate for a user’s facial image on a real-world smart-phone and the preference prediction accuracy of a collaborative filtering-based recommendation system. As a result, we confirmed that the system is highly effective and applicable to convergence by a location-based service and a content recommender through our implementation and preference prediction experiments.