• Journal of Internet Computing and Services
• /
• v.10 no.1
• /
• pp.111-121
• /
• 2009

In this paper, we propose a multimedia recommender system using user's playback time. Proposed system collects multimedia content which is requested by user and its user‘s playback time, as web log data. The system predicts playback time.based preference level and related contents from collected transaction database by fuzzy association rule mining. Proposed method has a merit which sorts recommendation list according to preference without user’s custom preference data, and prevents a false preference. As an experimental result, we confirm that proposed system discovers useful rules and applies them to recommender system from a transaction which doesn‘t include custom preferences.

• Journal of Digital Contents Society
• /
• v.11 no.1
• /
• pp.105-115
• /
• 2010

Search engines provide web documents that are related to user's query. However, using only the query terms that user provided, it is hard for search engines to know user's exact intention and provide the very matching web documents. To remedy this problem, search systems are needed to exploit personalized search technologies. In this paper, we propose not only a novel personalized query recommendation scheme based on folksonomy but also a new personalized search service architecture which reduces the risk of privacy violation while enabling search service providers to provide other various personalized services such as personalized advertisement.

• Journal of Intelligence and Information Systems
• /
• v.18 no.2
• /
• pp.131-141
• /
• 2012

In this paper, we propose a new method for collaborative filtering (CF)-based recommender systems. Traditional CF-based recommendation algorithms have applied constant settings such as a reference group (neighborhood) size and a significance level to all users. In this paper we develop a new method that identifies optimal personalized settings for each user and applies them to generating recommendations for individual users. Personalized parameters are identified through iterative simulations with 'training' and 'verification' datasets. The method is compared with traditional 'constant settings' methods using Netflix data. The results show that the new method outperforms traditional, ordinary CF. Implications and future research directions are also discussed.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.3
• /
• pp.189-199
• /
• 2009

The amount of information on the Web is explosively growing as the Internet gains in popularity. However, only a small portion of the information on the Web is truly relevant or useful to the user. Thus, offering suitable information according to user demand is an important subject in information retrieval. In e-commerce, the recommender system is essential to revitalize commercial transactions, raise user satisfaction and loyalty towards the information provider. The existing recommender systems are mostly based on user data collected at servers, so user data are dispersed over several servers. Therefore, web servers that lack sufficient user behavior data cannot easily infer user preferences. Also, if the user visits the server infrequently, it may be hard to reflect the dynamically changing user's interest. This paper proposes a novel personalization system analyzing the user preference based on web documents that are accessed by the user on a user device. The system also identifies non-content blocks appearing repeatedly in the dynamically generated web documents, and adds weight to the keywords extracted from the hyperlink sentence selected by the user. Therefore, the system establishes at an early stage recommendation strategies for the web server that has little user data. Also, user profiles are generated rapidly and more accurately by identifying the information blocks. In order to evaluate the proposed system, this study collected web data and purchase history from users who have current purchase activity. Then, we computed the similarity between purchase data and the user profile. We confirm the accuracy of the generated user profile since the web page containing the purchased item has higher correlation than other item pages.

• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.23-46
• /
• 2021

Collaborative filtering, which is often used in personalization recommendations, is recognized as a very useful technique to find similar customers and recommend products to them based on their purchase history. However, the traditional collaborative filtering technique has raised the question of having difficulty calculating the similarity for new customers or products due to the method of calculating similaritiesbased on direct connections and common features among customers. For this reason, a hybrid technique was designed to use content-based filtering techniques together. On the one hand, efforts have been made to solve these problems by applying the structural characteristics of social networks. This applies a method of indirectly calculating similarities through their similar customers placed between them. This means creating a customer's network based on purchasing data and calculating the similarity between the two based on the features of the network that indirectly connects the two customers within this network. Such similarity can be used as a measure to predict whether the target customer accepts recommendations. The centrality metrics of networks can be utilized for the calculation of these similarities. Different centrality metrics have important implications in that they may have different effects on recommended performance. In this study, furthermore, the effect of these centrality metrics on the performance of recommendation may vary depending on recommender algorithms. In addition, recommendation techniques using network analysis can be expected to contribute to increasing recommendation performance even if they apply not only to new customers or products but also to entire customers or products. By considering a customer's purchase of an item as a link generated between the customer and the item on the network, the prediction of user acceptance of recommendation is solved as a prediction of whether a new link will be created between them. As the classification models fit the purpose of solving the binary problem of whether the link is engaged or not, decision tree, k-nearest neighbors (KNN), logistic regression, artificial neural network, and support vector machine (SVM) are selected in the research. The data for performance evaluation used order data collected from an online shopping mall over four years and two months. Among them, the previous three years and eight months constitute social networks composed of and the experiment was conducted by organizing the data collected into the social network. The next four months' records were used to train and evaluate recommender models. Experiments with the centrality metrics applied to each model show that the recommendation acceptance rates of the centrality metrics are different for each algorithm at a meaningful level. In this work, we analyzed only four commonly used centrality metrics: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality. Eigenvector centrality records the lowest performance in all models except support vector machines. Closeness centrality and betweenness centrality show similar performance across all models. Degree centrality ranking moderate across overall models while betweenness centrality always ranking higher than degree centrality. Finally, closeness centrality is characterized by distinct differences in performance according to the model. It ranks first in logistic regression, artificial neural network, and decision tree withnumerically high performance. However, it only records very low rankings in support vector machine and K-neighborhood with low-performance levels. As the experiment results reveal, in a classification model, network centrality metrics over a subnetwork that connects the two nodes can effectively predict the connectivity between two nodes in a social network. Furthermore, each metric has a different performance depending on the classification model type. This result implies that choosing appropriate metrics for each algorithm can lead to achieving higher recommendation performance. In general, betweenness centrality can guarantee a high level of performance in any model. It would be possible to consider the introduction of proximity centrality to obtain higher performance for certain models.

• The Journal of Society for e-Business Studies
• /
• v.10 no.2
• /
• pp.109-132
• /
• 2005

With the flood of multimedia contents over the digital TV channels, the internet, and etc., users sometimes have a difficulty in finding their preferred contents, spend heavy surfing time to find them, and are even very likely to miss them while searching. In this paper we suggests two-step clustering technique using time schema on how the system can recommend the user's preferred contents based on the collaborative filtering that has been proved to be successful when new users appeared. This method maps and recommends users' profile according to the gender and age at the first step, and then recommends a probabilistic item clustering customers who choose the same item at the same time based on time schema at the second stage. In addition, this has improved the accuracy of predictions in recommendation and the efficiency in time calculation by reflecting feedbacks of the result of the recommender engine and dynamically update customers' preference.

• Journal of the Korean Society for Library and Information Science
• /
• v.45 no.4
• /
• pp.77-102
• /
• 2011

The purpose of this study is to design, implement and evaluate a personalized search system using gathered information on users to provide more accurate search results. For this purpose, a hybrid-based user profile is constructed by using subject classification. In order to evaluate the performance of the proposed system, experts directly measured and evaluated MRR, MAP and usability by using the Korean journal articles of science and technology DB. Its performance was better than the general search system in the area of "Computer Science" and "Library and Information Science". Especially better results were shown when tested on ambiguous keywords. Evaluation through in-depth interviews proved that the proposed personalized search system was more efficient in looking up and obtaining information. In addition, the proposed personalized search system provided a variety of recommendation systems which proved helpful in navigating for new information. High user satisfaction ratings on the proposed personalized search system were another proof of its usefulness. In this study, we were able to prove through expert evaluation that the proposed personalized search system was more efficient in information retrieval.

• Journal of The Korean Association of Information Education
• /
• v.17 no.4
• /
• pp.375-382
• /
• 2013

Collaborative filtering has been most widely used in commercial sites to recommend items based on the history of user preferences for items. The basic idea behind this method is to find similar users whose ratings for items are incorporated to make recommendations for new items. Hence, similarity calculation is most critical in recommendation performance. This paper presents a new similarity measure that takes each rating of a user relatively to his own ratings. Extensive experiments revealed that the proposed measure is more reliable than the classic measures in that it significantly decreases generation of extreme similarity values and its performance improves when consulting neighbors with high similarites only. In particular, the results show that the proposed measure is superior to the classic ones for datasets with large rating scales.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.5B
• /
• pp.542-551
• /
• 2009

We proposed the context based user profile which is aware of its user's situation and based on user's situation it recommends personalized services. The user profile which consists of (context, service) pair can be acquired by the context and the service usage of a user; it then can be used to recommend personalized services for the user. In this paper, we show how they can be evolved without previously known user information so that not to violate privacy during the learning phase; in the result our user profile can be applied to any new environment without any modification to model only except context profiles. Using context-awareness based user profile, the service usage pattern of a user can be learned by the union of contexts and the preferred services can be recommended by the current environments. Finally, we evaluate the precision of proposed approach using simulation with data sets of UCI depository and Weka tool-kit.

• The KIPS Transactions:PartD
• /
• v.17D no.6
• /
• pp.453-462
• /
• 2010

The switch from analog to digital broadcast television is extended rapidly. The DTV can offer multiple programming choices, interactive capabilities and so on. Moreover, with the spread of Internet, the information exchange between the communities is increasing, too. These facts lead to the new TV service environment which can offer customized TV programs to personal/community users. This paper proposes a 'Customized Digital TV System for Individuals/Communities based on Data Stream Mining' which can analyze user's pattern of TV watching behavior. Due to the characteristics of TV program data stream and EPG(electronic program guide), the data stream mining methods are employed in the proposed system. When a user is watching DTV, the proposed system can control the surrounding circumstances as using the user behavior profiles. Furthermore, the channel recommendation system on the smart phone environment is proposed to utilize the profiles widely.