• Annual Conference of KIPS
• /
• 2014.04a
• /
• pp.642-644
• /
• 2014

정보의 폭발적인 증가로 인해 사용자들은 오히려 원하는 정보를 빠른 시간에 얻는 것이 힘들어졌다. 따라서 이 문제를 해결하기 위한 다양한 방식의 새로운 서비스들이 제공되고 있다. 추천 시스템 중에서 영화를 추천해주는 방법에는 사용되는 알고리즘에는 협업필터링 방법이 가장 성공한 알고리즘으로 사용되고 있다. 협업 필터링 방법은 사용자가 자발적으로 입력한 선호도 평가치를 바탕으로 추천 하고자 하는 사용자와 취향이 비슷하다고 판단되는 사람들 즉, 최근접 이웃을 구하고 최근접 이웃의 선호도 평가치를 바탕으로 사용자에게 영화를 추천을 해주는 기법이다. 그러나 협업 필터링에는 몇 가지 대표적인 문제점이 있으며 희박성 및 확장성, 투명성이 있다. 본 논문에서는 영화 추천 시스템에서의 협업필터링의 희박성 문제를 보완하고자 개개인의 성향을 반영하여 효율이 좋은 추천 방법을 제안하고 하둡에서 성능평가를 하였다.

• Proceedings of the Korea Association of Information Systems Conference
• /
• 2002.11a
• /
• pp.333-347
• /
• 2002

• KAPE Magazine
• /
• s.227
• /
• pp.17-18
• /
• 2015

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.3
• /
• pp.147-155
• /
• 2018

In this paper, we tried to find out sound quality metrics to express discomfort of overload excavator noise and to develop sound quality indexes through multiple regression analysis by using them. For this purpose, the interior noise of cabin under overload condition was recorded for six excavator models with different noise properties and Jury test was carried out by PCM (Paired Comparison Method) and MEM (Magnitude Estimation Method). Jury test result with low consistency was classified into two groups with different preference tendencies by cluster analysis and multiple regression analysis was conducted in order to find out which sound quality metrics have significant effects on discomfort(low preference). As a result, we figured out that the sound quality metrics to express the discomfort were the partial loudness (= $PN_{10Bark}$) between 0 and 10 Bark in case of group1 and the difference between engine noise(= $dB_{EG}$) and hydraulic system noise ($dB_1$) in case of group2. Using the results of preference ranking and tendency analysis of PCM followed by the correlation analysis between PCM and MEM, the more reliable results were adopted by excluding the data with low consistency obtained from Jury test via MEM.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.10a
• /
• pp.3-8
• /
• 1998

본 논문에서는 사용자의 관심도나 선호도를 반영하여, 퍼지숫자를 정렬하는 방법을 제안한다. 사용자는 자신의 관심도나 선호도를 퍼지집합으로 표현한다. 제안하는 방법은 사용자로부터 주어진 퍼지집합을 평가관점으로 이용하며, 평가함수로는 이전에 제안된 만족도 함수를 이용한다. 제안하는 방법이 관점에 따라 어떠한 결과를 주는지를 보기 위하여, 퍼지숫자 정렬에 적용한 예를 보인다.

• Journal of the Korean Home Economics Association
• /
• v.40 no.9
• /
• pp.97-107
• /
• 2002

• Journal of the Korean Society for Library and Information Science
• /
• v.45 no.2
• /
• pp.163-184
• /
• 2011

The aim of this study is providing basic data to design successful in-service training program for teacher librarians by analyzing their preference on subjects of customized intensive programs for indicators of teacher expertise development. According to the survey, teacher librarians seem to regard training subjects such as Reading Education and Information Literacy Instruction related educational information services as core jobs and prefer developing instructional contents and materials. Under the levels of school it seems that teacher librarians in the elementary school are interested in programs for library activation, management of volunteers, analyses of users' needs and curricula, evaluations of user instruction and information literacy instruction. Older teacher librarians favor an understanding of metadata, building and supporting information system and instruction. Therefore, training subjects for teacher librarians should be formed by linking strategies between school library instruction and subject curricula. And in terms of the method of training, case studies and practical training might be better than lectures based on the theory.

• 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.

• The Journal of the Korea Contents Association
• /
• v.13 no.7
• /
• pp.379-385
• /
• 2013

On the internet, the rating scores assigned by customers are considered as the preference information of themselves and thus, these can be used efficiently in the customer profile generation process of recommender system. However, since anyone is free to assign a score that has a biased rating, using this without any filtering can exhibit a reliability problem. In this study, we suggest the methodology that measures the reliability of rating scores and then applies them to the customer profile creation process. Unlikely to some related studies which measure the reliability on the user level, we measure the reliability on the individual rating score level. Experimental results show that prediction accuracy of recommender system can be enhanced when ratings with higher reliability are selectively used for the customer profile configuration.

• Journal of KIISE:Software and Applications
• /
• v.29 no.9
• /
• pp.669-675
• /
• 2002

In recent years most of personalized recommender systems in electronic commerce utilize collaborative filtering algorithm in order to recommend more appropriate items. User-based collaborative filtering is based on the ratings of other users who have similar preferences to a user in order to predict the rating of an item that the user hasn't seen yet. This nay decrease the accuracy of prediction because the similarity between two users is computed with respect to the two users and only when an item has been rated by the users. In item-based collaborative filtering, the preference of an item is predicted based on the similarity between the item and each of other items that have rated by users. This method, however, uses the ratings of users who are not the neighbors of a user for computing the similarity between a pair of items. Hence item-based collaborative filtering may degrade the accuracy of a recommender system. In this paper, we present a new approach that a user's neighborhood is used when we compute the similarity between the items in traditional item-based collaborative filtering in order to compensate the weak points of the current item-based collaborative filtering and to improve the prediction accuracy. We empirically evaluate the accuracy of our approach to compare with several different collaborative filtering approaches using the EachMovie collaborative filtering data set. The experimental results show that our approach provides better quality in prediction and recommendation list than other collaborative filtering approaches.