• Journal of the Korean Operations Research and Management Science Society
• /
• v.31 no.2
• /
• pp.27-39
• /
• 2006

The Collaborative Filtering (CF) is one of the popular techniques for personalization in e-commerce storefronts. For CF-based recommendation, every customer needs to provide subjective evaluation ratings for some products based on his/her preference. Also, if an e-commerce site recommends a new product, some customers should rate it. However, there is no in-depth investigation on the impacts on recommendation performance of two number of ratings, i.e. the number of ratings of an individual customer and the number of ratings of an item, even though these are important factors to determine performance of CF methods. In this study, using publicly available EachMovie data set, we empirically investigate the relationships between the two number of ratings and the performance of CF. For the purpose, three analyses were executed. The first and second analyses were performed to investigate the relationship between the number of ratings of a particular customer and the recommendation performance of CF. In the third analysis, we investigate the relationship between the number of ratings on a particular item and the recommendation performance of CF. From these experiments, we can find that there are thresholds in terms of the number of ratings below which the recommendation performances increase monotonically. That is, the number of ratings of a customer and the number of ratings on an item are critical to the recommendation performance of CF when the number of ratings is less than the thresholds, but the value of the ratings decreases after the numbers of ratings pass the thresholds. The results of the experiments provide insight to making operational decisions concerning collaborative filtering in practice.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.4
• /
• pp.488-493
• /
• 2007

In general, recommendation systems quantify real-time context information obtained in the stage of collaborative filtering and use quantified context information in order to recommend services. But the recommendation systems can have problems of recommending inaccurate information because of lack of context information or classifying users into inaccurate groups because of simple classification works in the stage of quantification. In this paper, we solved the problems of lack of context information obtained in real-time by combining users' profile information used in the contents-based filtering and context information obtained in real-time. In addition, we tried collaborative filtering at the quantification stage by improving absolute classification methods to relative ones. As the result of experiments, this method improved prediction preference by 5.8% than real-time recommendation systems using context information in pure P2P environment.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.10
• /
• pp.3438-3457
• /
• 2022

As location-based services (LBS) are widely used in vehicular ad-hoc networks (VANETs), location privacy has become an utmost concern. Spatial cloaking is a popular location privacy protection approach, which uses a cloaking area containing k-1 collaborative vehicles (CVs) to replace the real location of the requested vehicle (RV). However, all CVs are assumed as honest in k-anonymity, and thus giving opportunities for dishonest CVs to submit false location information during the cloaking area construction. Attackers could exploit dishonest CVs' false location information to speculate the real location of RV. To suppress this threat, an edge-assisted Trusted Collaborative Anonymity construction scheme called TCA is proposed with trust mechanism. From the design idea of trusted observations within variable radius r, the trust value is not only utilized to select honest CVs to construct a cloaking area by restricting r's search range but also used to verify false location information from dishonest CVs. In order to obtain the variable radius r of searching CVs, a multiple linear regression model is established based on the privacy level and service quality of RV. By using the above approaches, the trust relationship among vehicles can be predicted, and the most suitable CVs can be selected according to RV's preference, so as to construct the trusted cloaking area. Moreover, to deal with the massive trust value calculation brought by large quantities of LBS requests, edge computing is employed during the trust evaluation. The performance analysis indicates that the malicious response of TCA is only 22% of the collaborative anonymity construction scheme without trust mechanism, and the location privacy leakage is about 32% of the traditional Enhanced Location Privacy Preserving (ELPP) scheme.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.11
• /
• pp.17-24
• /
• 2011

Collaborative filtering system is the clustering about user is built and then based on that clustering results will recommend the preferred item to the user. However, building user clustering is time consuming and also once the users evaluate and give feedback about the film then rebuilding the system is not simple. In this paper, genre pattern of movie recommendation systems is being used and in order to simplify and reduce time of rebuilding user clustering. A Frequent pattern networks is used and then extracts user preference genre patterns and through that extracted patterns user clustering will be built. Through built the clustering for all neighboring users to collaborative filtering is applied and then recommends movies to the user. When receiving user information feedback, traditional collaborative filtering is to rebuild the clustering for all neighbouring users to research and do the clustering. However by using frequent pattern Networks, through user clustering based on genre pattern, collaborative filtering is applied and when rebuilding user clustering inquiry limited by search time can be reduced. After receiving user information feedback through proposed user clustering based on genre pattern, the time that need to spent on re-establishing user clustering can be reduced and also enable the possibility of traditional collaborative filtering systems and recommendation of a similar performance.

• The Journal of the Korea Contents Association
• /
• v.10 no.7
• /
• pp.23-30
• /
• 2010

It is important for the strategy of cosmetic sales to investigate the sensibility and the preference degree in the environment that the makeup style has been changed focusing on the consumer center. We proposed the human sensibility ergonomics makeup recommendation system (MakeupRS) using the context sensor information applying the collaborative filtering technique as one of methods in the makeup style development centered on the consumer's sensibility and the preference. In the collaborative filtering technique, the Pearson correlation coefficient applying to the case amplification is used to calculate similarity weights between the users. To investigate the sensibility according to the effect of makeup styles, the makeup styles were analyzed in terms of 6 style factors, such as, the foundation, the color lens, the eye shadow, the eye lash, the cheek brusher, and the lipstick. Ultimately, this paper suggests empirical application to verify the adequacy and the validity with the human sensibility ergonomics makeup recommendation system.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.11
• /
• pp.133-140
• /
• 2012

In this paper, we propose a music recommendation method considering users' tags by collaborative tagging in a social music site. Since collaborative tagging allows a user to add keywords chosen by himself to web resources, it provides users' preference about the web resources concretely. In particular, emotional tags which represent human's emotion contain users' musical preference more directly than factual tags which represent facts such as musical genre and artists. Therefore, to classify the tags into the emotional tags and the factual tags and to assign weighted values to the emotional tags, a tag ontology called UniTag is developed. After preprocessing the tags, the weighted tags are used to create user profiles, and the music recommendation algorithm is executed based on the profiles. To evaluate the proposed method, a conventional playcount-based recommendation, an unweighted tag-based recommendation, and an weighted tag-based recommendation are executed. Our experimental results show that the weighted tag-based recommendation outperforms other two approaches in terms of precision.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.4 s.310
• /
• pp.50-57
• /
• 2006

In Electronic Commerce, the latest most of the personalized recommender systems have applied to the collaborative filtering technique. This method calculates the weight of similarity among users who have a similar preference degree in order to predict and recommend the item which hits to propensity of users. In this case, we commonly use Pearson Correlation Coefficient. However, this method is feasible to calculate a correlation if only there are the items that two users evaluated a preference degree in common. Accordingly, the accuracy of prediction falls. The weight of similarity can affect not only the case which predicts the item which hits to propensity of users, but also the performance of the personalized recommender system. In this study, we verify the improvement of the prediction accuracy through an experiment after observing the rule of the weight of similarity applying Vector similarity, Entropy, Inverse user frequency, and Default voting of Information Retrieval field. The result shows that the method combining the weight of similarity using the Entropy with Default voting got the most efficient performance.

• Journal of Korea Multimedia Society
• /
• v.13 no.3
• /
• pp.475-486
• /
• 2010

Contents Recommender System predicts user's preferences towards contents, and then recommends highly-predicted contents to user. Digital Identifier plays its part in identifying abstract works or digital contents in digital network environment. Digital Identifier could be effectively used in content-based filtering and collaborative filtering that are mainly used in Contents Recommender Systems. Therefore, this paper proposes an improvement of UCI metadata and resolution service for effective use of UCI in massive contents recommender systems. UCI metadata is expanded by adding elements such as abstract, keyword, genre, age, rate and review. Resolution service allows the operation systems to collect user preference for content by including input part of preference in a result page. This paper also designs and implements an improved UCI operation system and shows that the proposed improvement of UCI metadata and resolution service could be used for massive contents recommendation.

• The Journal of the Korea Contents Association
• /
• v.11 no.11
• /
• pp.1-9
• /
• 2011

With the development of next IT convergence technology and the construction of infrastructure for personalized healthcare services, the importance of services based on user's preference is being spotlighted. Healthcare service have been progressed as treatment and management for specific diseases and dietary nutrition managements to customers according to the increase in chronic patients. In this paper, we proposed the personalized dietary nutrition contents recommendation using the hybrid filtering for managing health. The proposed method uses the hybrid filtering through combining the collaborative filtering and the image filtering in order to reinforce the special trend that recommendation provides similar contents. We developed the Web application for this purpose, and experimented with it to verify the logical validity and effectiveness. Accordingly, the satisfaction and the quality of services will be improved the healthcare by recommending the dietary nutrition contents. This evaluation found that the difference of satisfaction by service was statistically meaningful and showed high satisfaction.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.6
• /
• pp.63-74
• /
• 2005

In Electronic Commerce, the latest most of the personalized recommender systems have applied to the collaborative filtering technique. This method calculates the weight of similarity among users who have a similar preference degree in order to predict and recommend the item which hits to propensity of users. In this case, we commonly use Pearson Correlation Coefficient. However, this method is feasible to calculate a correlation if only there are the items that two users evaluated a preference degree in common. Accordingly, the accuracy of prediction falls. The weight of similarity can affect not only the case which predicts the item which hits to propensity of users, but also the performance of the personalized recommender system. In this study, we verify the improvement of the prediction accuracy through an experiment after observing the rule of the weight of similarity applying Vector similarity, Entropy, Inverse user frequency, and Default voting of Information Retrieval field. The result shows that the method combining the weight of similarity using the Entropy with Default voting got the most efficient performance.