• Journal of Korea Multimedia Society
• /
• v.25 no.6
• /
• pp.819-832
• /
• 2022

Cross-domain recommender system transfers knowledge across different domains to improve the recommendation performance in a target domain that has a relatively sparse model. However, they suffer from the "negative transfer" in which transferred knowledge operates as noise. This paper proposes a novel Multiple Fusion-based Deep Cross-Domain Recommendation named MFDCR. We exploit Doc2Vec, one of the famous word embedding techniques, to fuse data user-wise and transfer knowledge across multi-domains. It alleviates the "negative transfer" problem. Additionally, we introduce a simple multi-layer perception to learn the user-item interactions and predict the possibility of preferring items by users. Extensive experiments with three domain datasets from one of the most famous services Amazon demonstrate that MFDCR outperforms recent single and cross-domain recommendation algorithms. Furthermore, experimental results show that MFDCR can address the problem of "negative transfer" and improve recommendation performance for multiple domains simultaneously. In addition, we show that our approach is efficient in extending toward more domains.

• Korean Management Science Review
• /
• v.21 no.3
• /
• pp.39-54
• /
• 2004

This paper presents a multi-agent system for intelligent recommendation of travel plans to users. The goal of the system is to provide alternative and preferable travel plans to users when the availability of tickets is low such as in vacations, holidays, weekends, or peak seasons. The multiple agents in the system search for available alternatives for a target travel in collaboration with other agents and recommend best alternatives by analyzing them using a multi-criteria decision-making model. A prototype online travel support system was constructed and a simulation experiment was performed for evaluation and comparison with different travel planning strategies.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.540-549
• /
• 2003

In collaborative recommendation two models are generally used: the user model and the article model. A user model learns correlation between users preferences and recommends an article based on other users preferences for the article. Similarly, an article model learns correlation between preferences for articles and recommends an article based on the target user's preference for other articles. In this paper, we investigates various combination methods of the user model and the article model for better recommendation performance. They include simple sequential and parallel methods, perceptron, multi-layer perceptron, fuzzy rules, and BKS. We adopt the multi-layer perceptron for training each of the user and article models. The multi-layer perceptron has several advantages over other methods such as the nearest neighbor method and the association rule method. It can learn weights between correlated items and it can handle easily both of symbolic and numeric data. The combined models outperform any of the basic models and our experiments show that the multi-layer perceptron is the most efficient combination method among them.

• Journal of Intelligence and Information Systems
• /
• v.18 no.3
• /
• pp.119-135
• /
• 2012

Among various recommendation techniques, neighborhood-based Collaborative Filtering (CF) techniques have been one of the most widely used and best performing techniques in literature and industry. This paper proposes new approaches that can enhance the neighborhood-based CF techniques by identifying a few best neighbors (the most similar users to a target user) more accurately with more information about neighbors. The proposed approaches put more weights to the users who have more items co-rated by the target user in similarity computation, which can help to better understand the preferences of neighbors and eventually improve the recommendation quality. Experiments using movie rating data empirically demonstrate simultaneous improvements in both recommendation accuracy and diversity. In addition to the typical single rating setting, the proposed approaches can be applied to the multi-criteria rating setting where users can provide more information about their preferences, resulting in further improvements in recommendation quality. We finally introduce a single metric that measures the balance between accuracy and diversity and discuss potential avenues for future work.

• Journal of KIISE:Software and Applications
• /
• v.31 no.4
• /
• pp.457-466
• /
• 2004

Recommendation is to offer information which fits user's interests and tastes to provide better services and to reduce information overload. It recently draws attention upon Internet users and information providers. The collaborative filtering is one of the widely used methods for recommendation. It recommends an item to a user based on the reference users' preferences for the target item or the target user's preferences for the reference items. In this paper, we propose a neural network based collaborative filtering method. Our method builds a model by learning correlation between users or items using a multi-layer perceptron. We also investigate integration of diverse information to solve the sparsity problem and selecting the reference users or items based on similarity to improve performance. We finally demonstrate that our method outperforms the existing methods through experiments using the EachMovie data.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2007.08b
• /
• pp.1508-1511
• /
• 2007

This paper introduces a new texture prediction for MVC( Multi-view Video Coding) which is currently being developed as an extension of the ITU-T Recommendation H.264 | ISO/IEC International Standard ISO/IEC 14496-10 AVC (Advanced Video Coding) [1]. The MVC's prcimary target is 3D video compression for 3D display system, thus, key technology compared to 2D video compression is reducing inter-view correlation. It is noticed, however, that the current JMVM [2] does not effectively eliminate inter-view correlation so that there is still a room to improve coding efficiency. The proposed method utilizes similarity of interview residual signal and can provide an additional coding gain. It is claimed that up to 0.2dB PSNR gain with 1.4% bit-rate saving is obtained for three multi-view test sequences.

• Journal of Intelligence and Information Systems
• /
• v.16 no.3
• /
• pp.181-199
• /
• 2010

As rapid progress of population aging and strong interest in health, the demand for new healthcare service is increasing. Until now healthcare service has provided post treatment by face-to-face manner. But according to related researches, proactive treatment is resulted to be more effective for preventing diseases. Particularly, the existing healthcare services have limitations in preventing and managing metabolic syndrome such a lifestyle disease, because the cause of metabolic syndrome is related to life habit. As the advent of ubiquitous technology, patients with the metabolic syndrome can improve life habit such as poor eating habits and physical inactivity without the constraints of time and space through u-healthcare service. Therefore, lots of researches for u-healthcare service focus on providing the personalized healthcare service for preventing and managing metabolic syndrome. For example, Kim et al.(2010) have proposed a healthcare model for providing the customized calories and rates of nutrition factors by analyzing the user's preference in foods. Lee et al.(2010) have suggested the customized diet recommendation service considering the basic information, vital signs, family history of diseases and food preferences to prevent and manage coronary heart disease. And, Kim and Han(2004) have demonstrated that the web-based nutrition counseling has effects on food intake and lipids of patients with hyperlipidemia. However, the existing researches for u-healthcare service focus on providing the predefined one-way u-healthcare service. Thus, users have a tendency to easily lose interest in improving life habit. To solve such a problem of u-healthcare service, this research suggests a u-healthcare recommender system which is based on collaborative filtering principle and social network. This research follows the principle of collaborative filtering, but preserves local networks (consisting of small group of similar neighbors) for target users to recommend context aware healthcare services. Our research is consisted of the following five steps. In the first step, user profile is created using the usage history data for improvement in life habit. And then, a set of users known as neighbors is formed by the degree of similarity between the users, which is calculated by Pearson correlation coefficient. In the second step, the target user obtains service information from his/her neighbors. In the third step, recommendation list of top-N service is generated for the target user. Making the list, we use the multi-filtering based on user's psychological context information and body mass index (BMI) information for the detailed recommendation. In the fourth step, the personal information, which is the history of the usage service, is updated when the target user uses the recommended service. In the final step, a social network is reformed to continually provide qualified recommendation. For example, the neighbors may be excluded from the social network if the target user doesn't like the recommendation list received from them. That is, this step updates each user's neighbors locally, so maintains the updated local neighbors always to give context aware recommendation in real time. The characteristics of our research as follows. First, we develop the u-healthcare recommender system for improving life habit such as poor eating habits and physical inactivity. Second, the proposed recommender system uses autonomous collaboration, which enables users to prevent dropping and not to lose user's interest in improving life habit. Third, the reformation of the social network is automated to maintain the quality of recommendation. Finally, this research has implemented a mobile prototype system using JAVA and Microsoft Access2007 to recommend the prescribed foods and exercises for chronic disease prevention, which are provided by A university medical center. This research intends to prevent diseases such as chronic illnesses and to improve user's lifestyle through providing context aware and personalized food and exercise services with the help of similar users'experience and knowledge. We expect that the user of this system can improve their life habit with the help of handheld mobile smart phone, because it uses autonomous collaboration to arouse interest in healthcare.

• Radiation Oncology Journal
• /
• v.14 no.4
• /
• pp.333-338
• /
• 1996

Purpose : To evaluate the clinical implications of scallop penumbra width that comes from multileaf collimator(MLC) effect by the daily routine patient setup error. Materials and Methods : The anales of $0^{circ},{\;}15^{circ},{\;}30^{circ},{\;}45^{circ},{\;}60^{circ},{\;}and{\;}75^{circ}$ inclined -radiation blocked fields were generated using the both conventional cerrobend block and the MLC. Film dosimetry in the phantom were performed to measure penumbral widths of differences between the dose distributions from the cerrobend block and those of respect the MLC. The patient setup error effect on scallop penumbra was simulated with respect to the table of setup error distribution. Same procedures are repeated for the cerrobend block generated field. Results : There are penumbral widths of to 3mm difference between the dose distributioins from two kinds of field shaping tools, the conventional block and the MLC with 4mm setup error model and resolution of 1cm leaf at the isocenter. Conclusion : We need not additive margin for MLC, if planning target volume is selected according to the recommendation of ICRU 50. For particular cases, we can include the target volume with less than 3mm additive margin.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2017.06a
• /
• pp.361-361
• /
• 2017

The NPK are known as macro elements that affect crop growth and yield. In 1989, Cambodia Agricultural Research and Development Institute (CARDI) gave a recommendation rate of fertilizer on rice production based on soil types. This recommended rate of NPK seems however relatively low as compared to farmers' practices nowadays and the amount in the neighboring countries. The CARDI recommended rate for Prateah Lang soil type is 50kg N, $25kg\;P_2O_5$, $25kg\;K_2O\;ha^{-1}$ while recent farmers' practice rates are 55 - 64kg N, 24 - 46kg $P_2O_5$, $30kg\;K_2O\;ha^{-1}$. However, the overuse of chemical fertilizer will lead to un-preferable plant growth, insect pest, disease and economic yield. Thus, we examined the effect of different NPK application rates on the growth and yield components in Prateah Lang soil type in Takeo province to investigate appropriate rates for improving rice productivity with economic efficiency. This study was conducted from July to November during wet season in 2013. A multi-locational trial with 6 treatments (T0 - T5) of NPK rates in 5 locations (trial 1 - 5) with 3 replications was conducted. The different combinations of NPK application were employed from 0, 50, 60, 80, 100, $120kg\;N\;ha^{-1}$, 0, 25, 30 45, $60kg\;P_2O_5\;ha^{-1}$ and 0, 15, 25, 30, $45kg\;K_2O\;ha^{-1}$. Urea, DAP and KCl were used for fertilization. Split application was employed [basal: 20% of N, 100% of P and K, top dressing-1st: 40% of N (30DAT), 2nd: 40% of N (PI stage)]. Three-week-old seedlings of var. Phka Rumdoul were transplanted with 2 - 3 seedlings $hill^{-1}$ with $20cm{\times}20cm$ spacing. Plant length, tiller number at the maximum tillering stage and yield components were measured. The different rates of NPK application affected some yield components. The panicle number per hill was the most important key component followed by the spikelet number per panicle. However, the other parameters such as the filled grain percentage and 1000 grains weight had small effect or weak relation with the yield. Although the panicle number per hill had a significantly positive correlation with the stem number per hill, it was not correlated with the percentage of productive culms. The variation in the grain yield among the 5 trials was small and the difference was not significant. Although the yield tended to be higher at higher N and P application, there was no significant difference above 60kg N and $30kg\;P_2O_5$. The yield was the highest at 15, 30 and $45kg\;K_2O$ followed by $25kg\;K_2O$. The relationships between N, P and the stem number per hill were significantly linear positive, though it was not linear between K and the stem number. From these results, to increase rice productivity in the target area, farmers' effort to increase N and P input rather than CARDI recommendation up to 60kg N and $30kg\;P_2O_5$ will be sufficient considering economic efficiency. Besides, the amount of K application should be reconsidered.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2017.06a
• /
• pp.363-363
• /
• 2017

The NPK are known as macro elements that affect crop growth and yield. In 1989, Cambodia Agricultural Research and Development Institute (CARDI) gave a recommendation rate of fertilizer on rice production based on soil types. This recommended rate of NPK seems however relatively low as compared to farmers' practices nowadays and the amount in the neighboring countries. The CARDI recommended rate for Prateah Lang soil type is 50kg N, 25kg P2O5, 25kg K2O ha-1 while recent farmers' practice rates are 55 - 64kg N, 24 - 46kg P2O5, 30kg K2O ha-1. However, the overuse of chemical fertilizer will lead to un-preferable plant growth, insect pest, disease and economic yield. Thus, we examined the effect of different NPK application rates on the growth and yield components in Prateah Lang soil type in Takeo province to investigate appropriate rates for improving rice productivity with economic efficiency. This study was conducted from July to November during wet season in 2013. A multi-locational trial with 6 treatments (T0 - T5) of NPK rates in 5 locations (trial 1 - 5) with 3 replications was conducted. The different combinations of NPK application were employed from 0, 50, 60, 80, 100, 120kg N ha-1, 0, 25, 30 45, 60kg P2O5 ha-1 and 0, 15, 25, 30, 45kg K2O ha-1. Urea, DAP and KCl were used for fertilization. Split application was employed [basal: 20% of N, 100% of P and K, top dressing-1st: 40% of N (30DAT), 2nd: 40% of N (PI stage)]. Three-week-old seedlings of var. Phka Rumdoul were transplanted with 2 - 3 seedlings hill-1 with $20cm{\times}20cm$ spacing. Plant length, tiller number at the maximum tillering stage and yield components were measured. The different rates of NPK application affected some yield components. The panicle number per hill was the most important key component followed by the spikelet number per panicle. However, the other parameters such as the filled grain percentage and 1000 grains weight had small effect or weak relation with the yield. Although the panicle number per hill had a significantly positive correlation with the stem number per hill, it was not correlated with the percentage of productive culms. The variation in the grain yield among the 5 trials was small and the difference was not significant. Although the yield tended to be higher at higher N and P application, there was no significant difference above 60kg N and 30kg P2O5. The yield was the highest at 15, 30 and 45kg K2O followed by 25kg K2O. The relationships between N, P and the stem number per hill were significantly linear positive, though it was not linear between K and the stem number. From these results, to increase rice productivity in the target area, farmers' effort to increase N and P input rather than CARDI recommendation up to 60kg N and 30kg P2O5 will be sufficient considering economic efficiency. Besides, the amount of K application should be reconsidered.