• Proceedings of the Korea Information Processing Society Conference
• /
• 한국정보처리학회 2013년도 춘계학술발표대회
• /
• pp.1067-1069
• /
• 2013

전자상거래 시장의 급격한 성장에 따라 고객이 원하는 정보를 얻기 위해 소요되는 시간과 노력을 절약하기 위한 방안으로 추천 시스템의 필요성이 강조되고 있다. 추천 시스템에 일반적으로 가장 많이 쓰이는 것이 협업필터링 기법이다. 협업 필터링은 추천시스템 분야에서 가장 성공적인 기법으로 전자상거래 포털에서 가장 널리 이용되고 있다. 그러나 희박성, 확장성, 투명성 등의 문제점을 가진다. 본 논문에서는 프로파일링 기법을 사용해 협업필터링의 희박성 문제 해소 방안으로 개인성향을 이용하여, 보다 정확한 추천을 하여 온라인 서점에 적용할 수 있는 추천 시스템이다.

• Journal of Information Management
• /
• 제33권1호
• /
• pp.1-17
• /
• 2002

This study examined several recommendation techniques to construct an effective book recommender system in a library. Experiments revealed that a hybrid recommendation technique is more effective than either collaborative filtering or content-based filtering technique in recommending books to be borrowed in an academic library setting. The recommendation technique based on association rule turned out the lowest in performance.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 한국방송∙미디어공학회 2020년도 하계학술대회
• /
• pp.587-590
• /
• 2020

최근, 디지털 콘텐츠 산업이 폭발적으로 성장됨에 따라 고객 유치를 위한 개인화 추천 기술들이 많은 주목을 받고 있다. 개인화 추천 방식들을 큰 갈래로 나누어 본다면 협업 필터링 기술과 내용 기반 기술로 나눌 수 있다. 협업 필터링의 경우 개인화 추천에는 적합하지만 사용자 평가 데이터의 양이 방대해야 하며 초기에 평가자가 없는 콘텐츠에 대해 추천할 수 없는 초기 평가자 문제가 존재한다. 따라서 매일 방대한 양의 콘텐츠가 편입되는 분야에서 사용하기에 큰 결점이 될 수 있다. 본 논문에서는 영화들의 정보가 담긴 데이터 셋과 사용자 평가 데이터, 그리고 사용자의 선호 기준을 의미하는 메타 가중치를 활용한 내용 기반의 맞춤형 영화 추천 시스템을 제안한다. 논문에서는 먼저, 영화를 고를 때 일반적으로 중요시 보는 속성들을 활용하여 영화의 특징 벡터를 구성하고, 이를 사용자 평가와 결합하여 개인의 선호에 대한 특징 벡터를 구성하는 방법을 제안하며, 구성된 데이터와 코사인 유사도, 메타 가중치를 활용하여 사용자 선호와 유사한 영화들을 도출하는 방법을 제안한다. 또한, 평가데이터를 활용하여 구현된 추천시스템의 검증 프로세스를 구성하고, 검증 프로세스를 활용한 손실 함수를 설계하여 적합한 메타 가중치를 학습하는 방법을 제시한다. 본 논문에서 제안하는 시스템은 다수의 속성을 조합하여 활용하므로 추천 결과가 과도하게 특수화 되지 않을 수 있으며, 메타 가중치라는 요소를 통해 더욱 개인화 된 추천을 제공할 수 있다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 한국컴퓨터정보학회 2015년도 제52차 하계학술대회논문집 23권2호
• /
• pp.169-172
• /
• 2015

웹툰의 양은 방대하여 사용자가 원하는 웹툰을 찾는데 어려움이 있기 때문에 체계적인 추천 시스템이 필요하다. 하지만 기존의 추천 시스템은 조회수가 많은 인기 웹툰을 추천하는 방식과 사용자와 비슷한 연령대, 성별의 사용자들이 조회한 콘텐츠를 추천해주는 인구 통계학적 추천(demographic filtering)방식, 그리고 비슷한 사용자를 분석하여 추천해주는 협업적 추천(collaborative filtering)방식에 국한되어 있어, 개인의 성향을 반영하여 추천하고 있다고 보기 어렵다. 따라서 사용자 개인의 성향을 분석하는 방식에 대한 시도가 필요하다. 본 연구에서는 이러한 한계를 극복하기 위해서 개인의 성향을 분석하는 지표로 MBTI(Myers-Briggs Type Indicator) 유형을 이용하고, 같은 MBTI 유형의 사용자간의 협업적 필터링 추천 방식을 제안하였다. 또, 협업적 필터링 방식에서 발생하는 콜드 스타트 문제와 초기 평가자 문제를 해결하는 방안을 제시하였다.

• Korean Business Review
• /
• 제20권2호
• /
• pp.105-118
• /
• 2007

Recommender system is the tool for E-commerce company based on the internet for increasing their sales ratio in the market. Recommender system suggests the list of items which night be wanted by customers. This list generated by the result of customers' preference prediction through the prediction algorithm automatically. Recommender system will be able to offer not only the important information for marketing strategy but also reduce the cost of customers' information retrieval trough the analysis of customers' purchase patterns and features. But there are several problems like as the extension of the users and items scales and if the recommendation to customers generated by unreliable recommender system makes the customer royalty to the system to weaken. In this study, we propose the criterion for pre-evaluation on the prediction performance only using the preference ratings on the items which are rated by customers before prediction process and we study the features of customers who are classified through this classification criterion.

• Journal of the Korean Data and Information Science Society
• /
• 제24권4호
• /
• pp.803-813
• /
• 2013

The purchase of items in e-commerce is a little bit different from that of items in off-line. The recommendation of items in off-line is conducted by salespersons' recommendation, However, the item recommendation in e-commerce cannot be recommended by salespersons, and so different types of methods can be recommended in e-commerce. Recommender system is a method which recommends items in e-commerce. Preferences of customers who want to purchase new items can be predicted by the preferences of customers purchasing existing items. In the recommender system, the items with estimated high preferences can be recommended to customers. The algorithm of collaborative filtering is used in recommender system of e-commerce, and the list of recommended items is made by estimated values, and then the list is recommended to customers. The dataset used in this research are 100k dataset and 1 million dataset in Movielens dataset. Similar results in two dataset are deducted for generalization. To suggest a new algorithm, distribution features of estimated values are analyzed by the existing algorithm and transformed algorithm. In addition, respondent'distribution features are analyzed respectively. To improve the collaborative filtering algorithm in neighborhood recommender system, a new algorithm method is suggested on the basis of existing algorithm and transformed algorithm.

• Journal of the Korean Institute of Intelligent Systems
• /
• 제19권2호
• /
• pp.242-247
• /
• 2009

Recently an intelligent system is developed for the service what users want not a passive system which just answered user's request. This intelligent system is used for personalized recommendation system and representative techniques are content-based and collaborative filtering. In this study, we propose a prediction system which is based on the techniques of recommendation system using a collaborative filtering and a fuzzy system to solve the collaborative filtering problems. In order to verify the prediction system, we used the data that is user's rating about movies. We predicted the user's rating using this data. The accuracy of this prediction system is determined by computing the RMSE(root mean square error) of the system's prediction against the actual rating about the each movie and is compared with the existing system. Thus, this prediction system can be applied to base technology of recommendation system and also recommendation of multimedia such as music and books.

• Journal of the Korean Data and Information Science Society
• /
• 제22권1호
• /
• pp.19-27
• /
• 2011

Recommendation systems are developed to overcome the problems of selection and to promote intention to use. In this study, we propose a recommendation system using adjacency data according to user's behavior over time. For this, the product adjacencies are identified from the adjacency matrix based on graph theory. This research finds that there is a trend in the users' behavior over time though product adjacency fluctuates over time. The system is tested on its usability. The tests show that implementing this recommendation system increases users' intention to purchase and reduces the search time.

• Journal of the Korean Institute of Intelligent Systems
• /
• 제21권5호
• /
• pp.609-617
• /
• 2011

This paper proposes a personalized food recommendation system using user profiles auto-generated from Twitter. The user profiles are generated by extracting nouns from Twitter, and calculating emotional scores according to whether each noun is collocated with emotion words. Representative noun information for each food is constructed by analyzing web pages relevant to foods. Appropriate foods for users can be recommended by calculating similarities among the extracted resources. The proposed system has an advantage in that it can always recommend foods even if a user is a newcomer.

• KIPS Transactions on Software and Data Engineering
• /
• 제11권11호
• /
• pp.447-454
• /
• 2022

In the era of the 4th industrial revolution, we are living in a flood of information. It is very difficult and complicated to find the information people need in such an environment. Therefore, in the flood of information, a recommendation system is essential. Among these recommendation systems, many studies have been conducted on each recommendation system for movies, music, food, and clothes. To date, most personalized recommendation systems have recommended clothes, books, or movies by checking individual tendencies such as age, genre, region, and gender. Future generations will want to be recommended clothes, books, and movies at once by checking age, genre, region, and gender. In this paper, we propose a recommendation system that recommends personalized clothes and food at once according to the user's emotions and weather. We obtained user data from Twitter of social media and analyzed this data as user's basic emotion according to Paul Eckman's theory. The basic emotions obtained in this way were converted into colors by applying Hayashi's Quantification Method III, and these colors were expressed as recommended clothes colors. Also, the type of clothing is recommended using the weather information of the visualcrossing.com API. In addition, various foods are recommended according to the contents of comfort food according to emotions.