• Journal of Information Technology Services
• /
• v.21 no.2
• /
• pp.109-126
• /
• 2022

The number and variety of products and services offered by companies have increased dramatically, providing customers with more choices to meet their needs. As a solution to this information overload problem, the provision of tailored services to individuals has become increasingly important, and the personalized recommender systems have been widely studied and used in both academia and industry. Existing recommender systems face important problems in practical applications. The most important problem is that it cannot clearly explain why it recommends these products. In recent years, some researchers have found that the explanation of recommender systems may be very useful. As a result, users are generally increasing conversion rates, satisfaction, and trust in the recommender system if it is explained why those particular items are recommended. Therefore, this study presents a methodology of providing an explanatory function of a recommender system using a review text left by a user. The basic idea is not to use all of the user's reviews, but to provide them in a summarized form using only reviews left by similar users or neighbors involved in recommending the item as an explanation when providing the recommended item to the user. To achieve this research goal, this study aims to provide a product recommendation list using user-based collaborative filtering techniques, combine reviews left by neighboring users with each product to build a model that combines text summary methods among deep learning-based natural language processing methods. Using the IMDb movie database, text reviews of all target user neighbors' movies are collected and summarized to present descriptions of recommended movies. There are several text summary methods, but this study aims to evaluate whether the review summary is well performed by training the Sequence-to-sequence+attention model, which is a representative generation summary method, and the BertSum model, which is an extraction summary model.

• Journal of Intelligence and Information Systems
• /
• v.14 no.1
• /
• pp.101-115
• /
• 2008

Combining collaborative filtering with some other technique is most common in hybrid recommender systems. As many recommended items from collaborative filtering seem to be similar with respect to content, the collaborative-content hybrid system suffers in terms of quality recommendation and recommending new items as well. To alleviate such problem, we have developed a novel method that uses a diversity metric to select the dissimilar items among the recommended items from collaborative filtering, which together with the input when fed into content space let us improve and include new items in the recommendation. We present experimental results on movielens dataset that shows how our approach performs better than simple content-based system and naive hybrid system.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.11
• /
• pp.475-482
• /
• 2013

Several approaches to recommendation systems have been studied. One of the most successful technologies for building personalization and recommendation systems is collaborative filtering, which is a technique that provides a process of filtering customer information based on such information profiles. Collaborative filtering systems, however, have a sparsity if there is not enough data to recommend. In this paper, we suggest a movie recommendation system, based on the weighted personal propensity and the collaborating filtering system, in order to provide a solution to such sparsity. Furthermore, we assess the system's applicability by using the open database MovieLens, and present a weighted personal propensity framework for improvement in the performance of recommender systems. We successfully come up with a movie recommendation system through the optimal personalization factors.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.11
• /
• pp.1116-1120
• /
• 2010

In this paper we present two refined neighbor selection algorithms for recommender systems and also show how the attributes of the items can be used for higher prediction quality. The refined neighbor selection algorithms adopt the transitivity-based neighbor selection method using virtual neighbors and alternate neighbors, respectively. The experimental results show that the recommender systems with the proposed algorithms outperform other systems and they can overcome the large scale dataset problem as well as the first rater problem without deteriorating prediction quality.

• The KIPS Transactions:PartD
• /
• v.14D no.3 s.113
• /
• pp.347-354
• /
• 2007

It is not easy for the customers to search the valuable information on the goods among countless items available in the Internet. In order to save time and efforts in searching the goods the customers want, it is very important for a recommender system to have a capability to predict accurately customers' preferences. In this paper we present a refined neighbor selection algorithm for clustering based collaborative filtering in recommender systems. The algorithm exploits a graph approach and searches more efficiently for set of influential customers with respect to a given customer; it searches with concepts of weighted similarity and ranked clustering. The experimental results show that the recommender systems using the proposed method find the proper neighbors and give a good prediction quality.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.1
• /
• pp.83-89
• /
• 2022

As a representative technique of recommender systems, collaborative filtering has been successfully in service through many commercial and academic systems. This technique recommends items highly rated by similar neighbor users, based on similarity of ratings on common items rated by two users. Recently research on time-aware recommender systems has been conducted, which attempts to improve system performance by reflecting user rating time of items. However, the decay rate uniform to past ratings has a risk of lowering the rating prediction performance of the system. This study proposes a rating time-aware similarity measure between users, which is a novel approach different from previous ones. The proposed approach considers changes of similarity value over time, not item rating time. In order to evaluate performance of the proposed method, experiments using various parameter values and types of time change functions are conducted, resulting in improving prediction accuracy of existing traditional similarity measures significantly.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.11a
• /
• pp.333-337
• /
• 2000

성공적인 사이트를 위한 필수적인 요소로 각광받고 있는 collaborative filtering 기술은 정보의 과부하를 줄일 수 있고 고객에 대한 충성도를 높여주는 효과로 인해 많은 사이트에 적용되어 운용되고 있다. 이 논문에서는 collaborative filtering 적용 포기에 발생하는 정보의 부족으로 인한 정확도 저하를 막기 위해 상품간 연관성을 이용한 결측티 예측 방안을 제안한다.

• Journal of Intelligence and Information Systems
• /
• v.13 no.4
• /
• pp.45-63
• /
• 2007

Recently as ubiquitous environment comes to the fore, information density is raised and enterprise is being able to capture and utilize customer-related information at the same time when the customer purchases a product. In this environment, a need for the recommender systems which can deliver proper information to the customer at the right time and right situation is highly increased. Therefore, the research on recommender systems continued actively in a variety of fields. Until now, most of recommender systems deal with item recommendation. However, in the market in ubiquitous environment where the same item can be purchased at several stores, it is highly desirable to recommend store to the customer based on his/her contextual situation and preference such as store location, store atmosphere, product quality and price, etc. In this line of research, we proposed the store recommender system using customer's contextual situation and preference in the market in ubiquitous environment. This system is based on collaborative filtering and Apriori algorithms. It will be able to provide customer-centric service to the customer, enhance shopping experiences and contribute in revitalizing market in the long term.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2007.11a
• /
• pp.602-609
• /
• 2007

Combining collaborative filtering with some other technique is most common in hybrid recommender systems. As many recommended items from collaborative filtering seem to be similar with respect to content, the collaborative-content hybrid system suffers in terms of quality recommendation and recommending new items as well. To alleviate such problem, we have developed a novel method that uses a diversity metric to select the dissimilar items among the recommended items from collaborative filtering, which together with the input when fed into content space let us improve and include new items in the recommendation. We present experimental results on movielens dataset that shows how our approach performs better than simple content-based system and naive hybrid system

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2006.06a
• /
• pp.344-353
• /
• 2006

Nowadays many people participate in online communities for information sharing. But most recommender systems are designed for personalization of individual user, so it is necessary to develop a recommendation procedure for group users, such as participants in online communities. This paper proposes a group recommender system to recommend books for group users in online communities. For such a purpose, we suggest a group recommendation procedure consisting of two phases. The first phase is to generate recommendation list for 'big user' using collaborative filtering, and the second phase is to remove irrelevant books among previous list reflecting the preference of each individual user. The procedure is explained step by step with an illustrative example. And this procedure can potentially be applied to other domains, such as music, movies and etc.