• Asia pacific journal of information systems
• /
• v.16 no.3
• /
• pp.159-176
• /
• 2006

Collaborative filtering is one of the popular techniques for personalized recommendation in e-commerce. In collaborative filtering, user profiles are usually managed per product category in order to reduce data sparsity. Product diversification of Internet storefronts and multiple product category sales of e-commerce portals require cross-product category usage of user profiles in order to overcome the cold start problem of collaborative filtering. In this paper, we study the feasibility of cross-product category usage of user profiles, and suggest a method to improve recommendation performance of cross-product category user profiling. First, we investigate whether user profiles on a product category can be used to recommend products in other product categories. Furthermore, a way of utilizing user profiles selectively is suggested to increase recommendation performance of cross-product category user profiling. The feasibility of cross-product category user profiling and the usefulness of the proposed method are tested with real click stream data of an Internet storefront which sells multiple product categories including books, music CDs, and DVDs. The experiment results show that user profiles on a product category can be used to recommend products in other product categories. Also, the selective usage of user profiles based on correlations between subcategories of two product categories provides better performance than the whole usage of user profiles.

• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1417-1424
• /
• 2018

In order to improve the performance of image classifications using Convolutional Neural Networks (CNN), applying a category hierarchy to the classification can be a useful idea. However, the visual separation of object categories is very different according to the upper and lower category levels and highly uneven in image classifications. Therefore, it is doubtable whether the use of category hierarchies for classification is effective in CNN. In this paper, we have clarified whether the image classification using category hierarchies improves classification performance, and found at which level of hierarchy classification is more effective. For experiments we divided the image classification task according to the upper and lower category levels and assigned image data to each CNN model. We identified and compared the results of three classification models and analyzed them. Through the experiments, we could confirm that classification effectiveness was not improved by reduction of number of categories in a classification model. And we found that only with the re-training method in the last network layer, the performance of lower category classification was not improved although that of higher category classification was improved.

• Korean Management Science Review
• /
• v.33 no.4
• /
• pp.1-15
• /
• 2016

Pairs trading is an arbitrage trading strategy using statistical properties of the spreads between two assets. This study analyzes the performance of the statistical pairs trading with the pairs selected from the same category as well as from the different category in the CME and other futures markets. Empirical results show that the pairs trading performance of the same category is poor whereas that of the different category proves profitable. This implies that the spreads between different category pairs can have the mean reversion property if pairs are properly selected using co-integration test, which is contrary to the existing research results on the overseas futures pairs trading.

• Proceeding of Spring/Autumn Annual Conference of KHA
• /
• 2008.11a
• /
• pp.479-482
• /
• 2008

The purpose of this study is a proposal about the flexibility of the performance category to revitalize the "Housing Performance Grading Indication System". This system consists of 5 main performance parts, 14 performance categories and 20 detailed performance lists. 5 main performance parts are Noise and Acoustics(Light-weight impact sound control, heavy-weight impact sound control, sound control of toilet, sound control of party wall), Long-life(flexibility, remodeling & maintenance, durability), Landscape & Indoor Environment(landscape, formaldehyde control& ventilation, daylighting, thermal environment), Welfare & Barrier-free(playground and community center, welfare space, barrier-free design), Fire Safety(fire safety, safe place, fire-resisting quality). The flexibility of the performance category is conformed to Long-life housing related grading and can realize the Long-life housing in response to resident needs about the space rearrangement. Now, The flexibility of the performance category is limited to the evaluation of the private space, that is, the Skeleton evaluation. This study would suggest alternatives to improve uppermost limits through analyzing 60 examples from January 9th 2009 to October 30th 2008. The flexibility of the performance category is possible to estimate not only the Skeleton evaluation but also the Infill evaluation in response to social changes.

• Journal of KIISE
• /
• v.44 no.9
• /
• pp.946-953
• /
• 2017

Semantic Role Decision defines the semantic relationship between the predicate and the arguments in natural language processing (NLP) tasks. The semantic role information and semantic category information should be used to make Semantic Role Decisions. The Sejong Electronic Dictionary contains frame information that is used to determine the semantic roles. In this paper, we propose a method to extend the Sejong electronic dictionary using word embedding and synonyms. The same experiment is performed using existing word-embedding and retrofitting vectors. The system performance of the semantic category assignment is 32.19%, and the system performance of the extended semantic category assignment is 51.14% for words that do not appear in the Sejong electronic dictionary of the word using the word embedding. The system performance of the semantic category assignment is 33.33%, and the system performance of the extended semantic category assignment is 53.88% for words that do not appear in the Sejong electronic dictionary of the vector using retrofitting. We also prove it is helpful to extend the semantic category word of the Sejong electronic dictionary by assigning the semantic categories to new words that do not have assigned semantic categories.

• Journal of the Korean Institute of Navigation
• /
• v.24 no.4
• /
• pp.227-234
• /
• 2000

In general, neural networks are widely used for the category classification of multispectral images. Since the input multispectral images into neural networks we, however, low contrast images, neural networks converge very slowly and are of bad performance. To overcome this problem, we propose a new image enhancement method which consists of smoothing process, finding the main valley and enhancement process. In addition the enhanced images by the proposed method are used as the input of neural networks for the category classification. When the new category classification method is applied to multispectral LANDSAT TM images, we verified that the neural networks converge very lastly and that the overall category classification performance is improved.

• Journal of Distribution Research
• /
• v.12 no.3
• /
• pp.23-48
• /
• 2007

Category management has been implemented to enhance competitiveness in the food distribution industry since 2000 in Korea. This study helps to understand why suppliers achieve better or worse performance than competitors in a category. The major objective of this article is to explore which category tactics are effective to have influence on category performance when suppliers as a category captain implement category management with variety enhancer categories like shampoo, toothpaste, and detergent. The Nielsen data were analyzed using regression and Chow test. The empirical results that were varied upon the store type and market position found out which specific actions on product assortments, pricing, shelving, and product replenishment can increase category sales. Specifically, in the case of market leader in large supermarket, the significant indicators of category sales with respect to category tactics are the out-of-stock rate, the variance across brand shares, the forward inventory, and the days supply of a product. However, in the case of follower in large supermarket, the significant indicators of category sales are the variance across brand shares, the forward inventory, and the days supply of a product. On the other hand, in the case of small supermarket, the significant factors on category sales for both market leader and follower are the retail distribution rate, the variance across brand shares, the forward inventory, and the days supply of a product category. In sum, regardless of the store type and market position, dominant brands in a category, the forward inventory, and short days supply of a product improved performance in all categories. Critical difference is that the out-of-stock rate acted as a key ingredient for the market leader between large and small supermarket and the retail distribution rate for the follower between large and small supermarket. This article presents some theoretical and managerial implications of the empirical results and finalizes the paper by addressing limitations and future research directions.

• Journal of Korean Society of Water and Wastewater
• /
• v.21 no.2
• /
• pp.165-175
• /
• 2007

Performance indicators haven't been used for the assessment of the wastewater treatment facility or management in Korea yet, therefore they are going to be important parts in wastewater utilities because they are used to understand present situation and to compare one with other wastewater utilities. In this study, we used performance indicators to assess the condition of wastewater utilities and they were divided into four categories (A, B, C, and D). A category represented the condition of the planning & construction and composed of wastewater supply, disaster defence and employees. B category represented maintenance of wastewater utilities and were composed of manhole, sewer, and technical employees. C category showed the operation efficiency of wastewater utilities and D category represented the environmental load. To analyze the situation of wastewater utilities overall, cluster analysis was performed using four categori' es indicators. And CCC (Cubic Clustering Criterion) and R-square were used to decide the proper number of clusters, and wastewater utilities of 48 cities were divided into 5 groups(I, II, III, IV, and V groups). Each cluster was analyzed by average and standard deviation to understand the situation of wastewater utilities. A group analysis showed that IV and V clusters were insufficient, B group showed that I and IV groups were insufficient, C group showed all clusters are above average, and D group was also like C group.

• Journal of the Korean Society for information Management
• /
• v.25 no.1
• /
• pp.211-233
• /
• 2008

This study examines various weighting methods for improving the performance of automatic classification based on Rocchio algorithm on two collections(LISA, Reuters-21578). First, three factors for weighting are identified as document factor, document factor, category factor for each weighting schemes, the performance of each was investigated. Second, the performance of combined weighting methods between the single schemes were examined. As a result, for the single schemes based on each factor, category-factor-based schemes showed the best performance, document set-factor-based schemes the second, and document-factor-based schemes the worst. For the combined weighting schemes, the schemes(idf*cat) which combine document set factor with category factor show better performance than the combined schemes(tf*cat or ltf*cat) which combine document factor with category factor as well as the common schemes (tfidf or ltfidf) that combining document factor with document set factor. However, according to the results of comparing the single weighting schemes with combined weighting schemes in the view of the collections, while category-factor-based schemes(cat only) perform best on LISA, the combined schemes(idf*cat) which combine document set factor with category factor showed best performance on the Reuters-21578. Therefore for the practical application of the weighting methods, it needs careful consideration of the categories in a collection for automatic classification.

• The Journal of Asian Finance, Economics and Business
• /
• v.8 no.1
• /
• pp.741-750
• /
• 2021

The research aims to analyze the level of efficiency by grouping banks during the period 2017 - 2018 into category 1 and category 2 banks and then dividing them as Regional Development Banks (BPD) and Non-BPD Conventional Commercial Banks (BUK) within each category. The research objects are banks within the categories BPD and BUK comprised 18 BPDs and 35 BUKs. The research methodology uses 3 stages, first, using Data Envelopment Analysis (DEA) we measure the level of bank efficiency; second, using the Tobit regression model we evaluate the effect of financial performance on DEA efficiency, and third, using the Mann-Whitney test we determine whether there is a difference in the efficiency of category 1 and 2 banks. The results showed that there was a decrease in the efficiency of category 1 and 2 banks but on average, the efficiency of category 1 banks is higher than category 2 banks. The estimation results of the Tobit regression model show that only the ROA variable affects the efficiency level of category 1 banks, while category 2 banks are influenced by NPL and ROA variables. In the Mann-Whitney test, it was proven that there were differences in efficiency between BUK and BPD in category 1 and 2 banks.