• Journal of the Korea Society of Computer and Information
• /
• v.25 no.5
• /
• pp.99-106
• /
• 2020

The scalability problem inherent in collaborative filtering-based recommender systems has been an issue in related studies during past decades. Clustering is a well-known technique for handling this problem, but has not been actively studied due to its low performance. This paper adopts a clustering method to overcome the scalability problem, inherent drawback of collaborative filtering systems. Furthermore, in order to handle performance degradation caused by applying clustering into collaborative filtering, we take two strategies into account. First, we use fuzzy clustering and secondly, we propose and apply a similarity estimation method based on user preference for movie genres. The proposed method of this study is evaluated through experiments and compared with several previous relevant methods in terms of major performance metrics. Experimental results show that the proposed demonstrated superior performance in prediction and rank accuracies and comparable performance to the best method in our experiments in recommendation accuracy.

• Journal of Information Science Theory and Practice
• /
• v.9 no.2
• /
• pp.66-82
• /
• 2021

Recommender Systems have gained immense popularity due to their capability of dealing with a massive amount of information in various domains. They are considered information filtering systems that make predictions or recommendations to users based on their interests and preferences. The more recent technology, Linked Open Data (LOD), has been introduced, and a vast amount of Resource Description Framework data have been published in freely accessible datasets. These datasets are connected to form the so-called LOD cloud. The need for semantic data representation has been identified as one of the next challenges in Recommender Systems. In a LOD-enabled recommendation framework where domain awareness plays a key role, the semantic information provided in the LOD can be exploited. However, dealing with a big chunk of the data from the LOD cloud and its integration with any domain datasets remains a challenge due to various issues, such as resource constraints and broken links. This paper presents the challenges of interconnecting and extracting the DBpedia data with the MovieLens 1 Million dataset. This study demonstrates how LOD can be a vital yet rich source of content knowledge that helps recommender systems address the issues of data sparsity and insufficient content analysis. Based on the challenges, we proposed a few alternatives and solutions to some of the challenges.

• Journal of The Korean Association of Information Education
• /
• v.25 no.6
• /
• pp.927-934
• /
• 2021

AI(Artificial intelligence) is gradually taking up a large part of our lives, and the pace of AI development is accelerating. It is called ACT that develop students' computational thinking in the way artificial intelligence learns. Among ACTs, pattern recognition is an essential factor in efficiently solving problems. Pattern analysis is part of the pattern recognition process. In fact, Netflix's personalized movie recommendation service and what it named Covid-19 after repeated symptoms are all the results of pattern analysis. While the importance of ACT, including pattern recognition, is highlighted, software education for kindergarten and elementary school lower grades is much insufficient compared to foreign countries. Therefore, this study aims to design and develop textbooks for the development of artificial intelligence-based computational thinking through pattern analysis for kindergarten students.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.19-38
• /
• 2017

Collaborative filtering (CF) algorithm has been popularly used for implementing recommender systems. Until now, there have been many prior studies to improve the accuracy of CF. Among them, some recent studies adopt 'hybrid recommendation approach', which enhances the performance of conventional CF by using additional information. In this research, we propose a new hybrid recommender system which fuses CF and the results from the social network analysis on trust and distrust relationship networks among users to enhance prediction accuracy. The proposed algorithm of our study is based on memory-based CF. But, when calculating the similarity between users in CF, our proposed algorithm considers not only the correlation of the users' numeric rating patterns, but also the users' in-degree centrality values derived from trust and distrust relationship networks. In specific, it is designed to amplify the similarity between a target user and his or her neighbor when the neighbor has higher in-degree centrality in the trust relationship network. Also, it attenuates the similarity between a target user and his or her neighbor when the neighbor has higher in-degree centrality in the distrust relationship network. Our proposed algorithm considers four (4) types of user relationships - direct trust, indirect trust, direct distrust, and indirect distrust - in total. And, it uses four adjusting coefficients, which adjusts the level of amplification / attenuation for in-degree centrality values derived from direct / indirect trust and distrust relationship networks. To determine optimal adjusting coefficients, genetic algorithms (GA) has been adopted. Under this background, we named our proposed algorithm as SNACF-GA (Social Network Analysis - based CF using GA). To validate the performance of the SNACF-GA, we used a real-world data set which is called 'Extended Epinions dataset' provided by 'trustlet.org'. It is the data set contains user responses (rating scores and reviews) after purchasing specific items (e.g. car, movie, music, book) as well as trust / distrust relationship information indicating whom to trust or distrust between users. The experimental system was basically developed using Microsoft Visual Basic for Applications (VBA), but we also used UCINET 6 for calculating the in-degree centrality of trust / distrust relationship networks. In addition, we used Palisade Software's Evolver, which is a commercial software implements genetic algorithm. To examine the effectiveness of our proposed system more precisely, we adopted two comparison models. The first comparison model is conventional CF. It only uses users' explicit numeric ratings when calculating the similarities between users. That is, it does not consider trust / distrust relationship between users at all. The second comparison model is SNACF (Social Network Analysis - based CF). SNACF differs from the proposed algorithm SNACF-GA in that it considers only direct trust / distrust relationships. It also does not use GA optimization. The performances of the proposed algorithm and comparison models were evaluated by using average MAE (mean absolute error). Experimental result showed that the optimal adjusting coefficients for direct trust, indirect trust, direct distrust, indirect distrust were 0, 1.4287, 1.5, 0.4615 each. This implies that distrust relationships between users are more important than trust ones in recommender systems. From the perspective of recommendation accuracy, SNACF-GA (Avg. MAE = 0.111943), the proposed algorithm which reflects both direct and indirect trust / distrust relationships information, was found to greatly outperform a conventional CF (Avg. MAE = 0.112638). Also, the algorithm showed better recommendation accuracy than the SNACF (Avg. MAE = 0.112209). To confirm whether these differences are statistically significant or not, we applied paired samples t-test. The results from the paired samples t-test presented that the difference between SNACF-GA and conventional CF was statistical significant at the 1% significance level, and the difference between SNACF-GA and SNACF was statistical significant at the 5%. Our study found that the trust/distrust relationship can be important information for improving performance of recommendation algorithms. Especially, distrust relationship information was found to have a greater impact on the performance improvement of CF. This implies that we need to have more attention on distrust (negative) relationships rather than trust (positive) ones when tracking and managing social relationships between users.

• Journal of Intelligence and Information Systems
• /
• v.18 no.4
• /
• pp.19-42
• /
• 2012

Consumer consumption patterns are shifting rapidly as buyers migrate from offline markets to e-commerce routes, such as shopping channels on TV and internet shopping malls. In the offline markets consumers go shopping, see the shopping items, and choose from them. Recently consumers tend towards buying at shopping sites free from time and place. However, as e-commerce markets continue to expand, customers are complaining that it is becoming a bigger hassle to shop online. In the online shopping, shoppers have very limited information on the products. The delivered products can be different from what they have wanted. This case results to purchase cancellation. Because these things happen frequently, they are likely to refer to the consumer reviews and companies should be concerned about consumer's voice. E-commerce is a very important marketing tool for suppliers. It can recommend products to customers and connect them directly with suppliers with just a click of a button. The recommender system is being studied in various ways. Some of the more prominent ones include recommendation based on best-seller and demographics, contents filtering, and collaborative filtering. However, these systems all share two weaknesses : they cannot recommend products to consumers on a personal level, and they cannot recommend products to new consumers with no buying history. To fix these problems, we can use the information which has been collected from the questionnaires about their demographics and preference ratings. But, consumers feel these questionnaires are a burden and are unlikely to provide correct information. This study investigates combining collaborative filtering with the centrality of social network analysis. This centrality measure provides the information to infer the preference of new consumers from the shopping history of existing and previous ones. While the past researches had focused on the existing consumers with similar shopping patterns, this study tried to improve the accuracy of recommendation with all shopping information, which included not only similar shopping patterns but also dissimilar ones. Data used in this study, Movie Lens' data, was made by Group Lens research Project Team at University of Minnesota to recommend movies with a collaborative filtering technique. This data was built from the questionnaires of 943 respondents which gave the information on the preference ratings on 1,684 movies. Total data of 100,000 was organized by time, with initial data of 50,000 being existing customers and the latter 50,000 being new customers. The proposed recommender system consists of three systems : [+] group recommender system, [-] group recommender system, and integrated recommender system. [+] group recommender system looks at customers with similar buying patterns as 'neighbors', whereas [-] group recommender system looks at customers with opposite buying patterns as 'contraries'. Integrated recommender system uses both of the aforementioned recommender systems to recommend movies that both recommender systems pick. The study of three systems allows us to find the most suitable recommender system that will optimize accuracy and customer satisfaction. Our analysis showed that integrated recommender system is the best solution among the three systems studied, followed by [-] group recommended system and [+] group recommender system. This result conforms to the intuition that the accuracy of recommendation can be improved using all the relevant information. We provided contour maps and graphs to easily compare the accuracy of each recommender system. Although we saw improvement on accuracy with the integrated recommender system, we must remember that this research is based on static data with no live customers. In other words, consumers did not see the movies actually recommended from the system. Also, this recommendation system may not work well with products other than movies. Thus, it is important to note that recommendation systems need particular calibration for specific product/customer types.

• The Journal of Society for e-Business Studies
• /
• v.17 no.3
• /
• pp.85-103
• /
• 2012

This study focused on user evaluation for personalized recommender systems with the integrated view of performance of the system and user attitude of recommender systems. Since users' evaluations of recommender systems can be affected by recommendation outcomes and presentation methods, both system performances based on outcomes and user attitudes formed by the presentation methods should be considered when explaining users' evaluations. However, an integrated view of system performance and user attitudes has not been applied to explain users' evaluation of recommender systems. Thus, the goal of this study is to explain users' evaluations of recommender systems under the integrated view of predictive features and explanation features at the same time. Our findings suggest that social presence, both accuracy and noveltyhave impacts onuser satisfaction for recommender systems. Especially, predictive features including accuracy and novelty affected user satisfaction. Novelty as well as accuracy is one of the significant factors for user satisfaction while recommender systems provided usual items users have experienced when systems provide serendipitous items. Likewise, explanation features with social presence and self-reference were important for user evaluation of personalized recommender systems. For explanation features, while social presence appears as one of important factors to user satisfaction of evaluating personalized recommendations, self-reference has no significant effect on user's satisfaction for recommender systems when compared to the result of social presence. Self-referencing messages did not affect user satisfaction but the levels of self-referencing are different between low and high groups in the experiment.

• Journal of Korean Elementary Science Education
• /
• v.21 no.2
• /
• pp.263-288
• /
• 2002

'Science Garden' is a science program that develops acquirements of science & technology at an early stage through experiential study and inspire the spirit of scientific inquiry by carrying out laboratory work, science movie, science lecture, scientific work, outdoor activities, computer classes and science experiment which is hard to do through the regular educational course. It is targeted grade 4,5,6 students during summer and winter vacation. 'Science Garden' is conducted by selective participation freely. It is a wholelistic activity that develops children's potential talents or creativities, improves interest and attitude toward science, and also gives opportunity for self-realization by extending capacity for inquiry to show each student's ability. This study is observed and is compared how it is conducted in elementary school using qualitative study. This study is used narrative observation, in-depth interview and document analysis. Objects of narrative observation are two elementary schools, each from Jellabukdo and Gyeonggido, and 7 teachers were interviewed in-depthly. Here are results of the study. 1 A teacher in school G never takes part in student activities, tends to be indifferent to classes, but focuses on observation and experiment in laboratory. And feedback or evaluation about student's activity is never done. On the other hand, a teacher in school S guides students to understand the principals of science on themselves, and wide variety forms of activities such as role playing, discussion, and games are being done. But an effort to evaluate student's activity is not being made properly. 2. Teachers set a high valuation on the need of 'Science Garden'. Observing the way of conducting 'Science Garden', usually teachers who is in charge of science for official work is selected as a teacher in charge, and groan under a heavy burden of conducting it without anyone's help. Participating students are selected by volunteering or teacher's recommendation, but because of low Participation rate, teachers have difficulty in conducting it. Plan for conducting ‘science Garden’ is made 20 days before it, after getting an official document from Office of Education, refering to booklets produced by National Jungang Science Institute, or data from Office of Education, and internet. Teachers evaluate rarely Most school principals have interest in 'Science Garden' but parents are not well aware of it, The budget is made at the same time with the plan, and scale is varied between 200,000∼500,000 won. Because of the improper way of selecting teacher in charge, difficulty in selecting students, heavy works caused by planning, conducting the program, as well as reporting teacher's work, and lack of parents' awareness, 'Science Garden' has been conducted formally and superficially. Next gives you direction to change, for the right way of conducting' Science Garden'. It is important to motivate competent teachers to instruct students actively. For students' active participation, They should publicize thoroughly beforehand, and develope 'Science Garden' program for teachers to be able to make better use of it. Evaluation of student activities and program should be done in the aspect of developing students’ faculties. Beside of school facilities, they need to put diverse local facilities and places to practical use for immediate natural experience. And not only separate schools but also associated form of schools to conduct it is necessary.

• 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.

• Journal of Intelligence and Information Systems
• /
• v.20 no.2
• /
• pp.137-148
• /
• 2014

Recommender system has become one of the most important technologies in e-commerce in these days. The ultimate reason to shop online, for many consumers, is to reduce the efforts for information search and purchase. Recommender system is a key technology to serve these needs. Many of the past studies about recommender systems have been devoted to developing and improving recommendation algorithms and collaborative filtering (CF) is known to be the most successful one. Despite its success, however, CF has several shortcomings such as cold-start, sparsity, gray sheep problems. In order to be able to generate recommendations, ordinary CF algorithms require evaluations or preference information directly from users. For new users who do not have any evaluations or preference information, therefore, CF cannot come up with recommendations (Cold-star problem). As the numbers of products and customers increase, the scale of the data increases exponentially and most of the data cells are empty. This sparse dataset makes computation for recommendation extremely hard (Sparsity problem). Since CF is based on the assumption that there are groups of users sharing common preferences or tastes, CF becomes inaccurate if there are many users with rare and unique tastes (Gray sheep problem). This study proposes a new algorithm that utilizes Social Network Analysis (SNA) techniques to resolve the gray sheep problem. We utilize 'degree centrality' in SNA to identify users with unique preferences (gray sheep). Degree centrality in SNA refers to the number of direct links to and from a node. In a network of users who are connected through common preferences or tastes, those with unique tastes have fewer links to other users (nodes) and they are isolated from other users. Therefore, gray sheep can be identified by calculating degree centrality of each node. We divide the dataset into two, gray sheep and others, based on the degree centrality of the users. Then, different similarity measures and recommendation methods are applied to these two datasets. More detail algorithm is as follows: Step 1: Convert the initial data which is a two-mode network (user to item) into an one-mode network (user to user). Step 2: Calculate degree centrality of each node and separate those nodes having degree centrality values lower than the pre-set threshold. The threshold value is determined by simulations such that the accuracy of CF for the remaining dataset is maximized. Step 3: Ordinary CF algorithm is applied to the remaining dataset. Step 4: Since the separated dataset consist of users with unique tastes, an ordinary CF algorithm cannot generate recommendations for them. A 'popular item' method is used to generate recommendations for these users. The F measures of the two datasets are weighted by the numbers of nodes and summed to be used as the final performance metric. In order to test performance improvement by this new algorithm, an empirical study was conducted using a publically available dataset - the MovieLens data by GroupLens research team. We used 100,000 evaluations by 943 users on 1,682 movies. The proposed algorithm was compared with an ordinary CF algorithm utilizing 'Best-N-neighbors' and 'Cosine' similarity method. The empirical results show that F measure was improved about 11% on average when the proposed algorithm was used