• Proceedings of the CALSEC Conference
• /
• 2004.02a
• /
• pp.239-250
• /
• 2004

The preference of customers change as time goes by. The existing Collaborative Filtering (CF) techniques has no room for including this change yet, although these techniques have been known to be the most successful recommendation technique that has been used in a number of different applications. In this study, we proposed a new methodology for enhancing the quality of recommendation using the customers' dynamic behaviors over time. The proposed methodology is applied to a large department store in Korea, compared to existing CF techniques. Some experiments on the real world data show that the proposed methodology provides higher quality recommendations than other CF techniques, especially better performance on heavy users.

• Journal of Korean Institute of Industrial Engineers
• /
• v.29 no.1
• /
• pp.90-99
• /
• 2003

Object clustering, which makes classification for a set of objects into a number of groups such that objects included in a group have similar characteristic and objects in different groups have dissimilar characteristic each other, has been exploited in diverse area such as information retrieval, data mining, group technology, etc. In this study, an object-clustering problem with similarity coefficients between objects is considered. At first, an evaluation function for the optimization problem is defined. Then, a genetic algorithm and local optimization technique based on heuristic method are proposed and used in order to obtain near optimal solutions. Solutions from the genetic algorithm are improved by local optimization techniques based on object relocation and cluster merging. Throughout extensive experiments, the validity and effectiveness of the proposed algorithms are tested.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10b
• /
• pp.69-71
• /
• 2000

군집화는 데이터 집합을 유사한 데이터 개체들의 군집들로 분할하여 데이터 속에 존재하는 의미 있는 정보를 얻는 과정이다. 대부분의 군집화 기법들은 비교적 적은 양의 데이터를 대상으로 한 것이고 다차원 대용량의 데이터 처리에 관한 문제는 다루지 않고 있어서 데이터 마이닝을 위한 군집화 기법으로는 부적절하다. 따라서 본 논문을 통해 대용량의 데이터에 적용할 수 있는 새로운 군집화 알고리즘인 계층적 대표값 군집화(HRC) 기법을 제안한다. HRC는 자기조직화지도와 계층적 군집화 기법을 접목한 하이브리드 방법으로 두 단계에 거쳐 군집화를 수행한다. 첫 번째 단계에서 자기조직화지도를 통해 데이터를 요약하고, 두 번째 단계에서 요약된 대표값 정보만을 가지고 계층적인 군집화를 수행한다. 또한, 두 번째 단계의 계층적 군집화 적용시 양질의 군집을 발견하기 위해 군집간의 유사도를 측정하는 새로운 척도를 고안하였다. 그리고 실험을 통해 HRC와 기존 군집화 알고리즘이 발견한 군집의 질을 비교하여 성능을 평가했다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.04a
• /
• pp.730-733
• /
• 2015

대용량의 인터넷 뉴스 데이터로부터 유용한 정보를 찾기 위해 연관 키워드, 핫 키워드 분석과 같은 다양한 분석 기술들이 연구되고 있다. 기존의 토픽 모델 기반의 기법은 키워드들간의 연관성을 제대로 표현하지 못하여 마이닝한 연관 키워드와 핫 키워드의 정확도가 낮은 문제점이 있다. 최근, 뉴스 데이터를 뉴스 내의 단어를 버텍스로, 같은 문장내의 단어들을 에지로 연결하는 그래프 기반의 모델링기법이 연구되었다. 이러한 뉴스 그래프 DB에서 그래프 마이닝 기술을 적용하면 연관 키워드, 핫 키워드를 마이닝 할 수 있다. 본 논문은 그래프 마이닝 기술 기반의 효과적인 뉴스 데이터 분석 기술을 제안한다. 실제 뉴스 데이터를 통해 마이닝한 유용한 뉴스 그래프 패턴들을 보이고 뉴스 데이터 분석에 효과적으로 활용될 수 있음을 보인다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.15 no.3
• /
• pp.159-165
• /
• 2015

The need for the recognition of music emotion has become apparent in many music information retrieval applications. In addition to the large pool of techniques that have already been developed in machine learning and data mining, various emerging applications have led to a wealth of newly proposed techniques. In the music information retrieval community, many studies and applications have concentrated on tag-based music recommendation. The limitation of music emotion tags is the ambiguity caused by a single music tag covering too many subcategories. To overcome this, multiple tags can be used simultaneously to specify music clips more precisely. In this paper, we propose a novel technique to rank the proper tag combinations based on the acoustic similarity of music clips.

• The Journal of Fisheries Business Administration
• /
• v.52 no.1
• /
• pp.23-46
• /
• 2021

The purpose of this study is to understand the production efficiency of individual fishing communities and provide directions for improvement. The subject of the study is aquaculture type Ochon-Gye in Goheung-gun. The analysis method used bootstrap-DEA to overcome the statistical reliability problem of the traditional DEA analysis technique. In addition, data mining-GIS was applied to identify the spatial productivity of fishing communities. The values of technology efficiency, pure technology efficiency, and scale efficiency were estimated for 32 aquaculture-type fishing villages. Then, using the benchmarking reference set and weights, the projection was presented through adjustment of the input factor excess, and furthermore, the confidence interval of the efficiency values considering statistical significance was estimated using bootstrap.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.429-430
• /
• 2019

본 논문에서는 최적화 기법에 기반한 지능형 시스템의 재무응용사례를 다룬다. 본 연구에서 제안하는 모형은 대표적인 최적화 기법 중 하나인 시뮬레이티드 어니일링인데 이는 유전자 알고리듬과 유사한 최적화 성능을 가지고 있는 것으로 알려져 있으나 재무분야에서 응용된 사례가 거의 없다. 본 연구에서 제안하는 지능형 시스템은 시뮬레이티드 어니일링과 기계학습 기법을 결합한 것이다. 일반적으로 최적화와 기계학습 기법을 결합하는 방법은 특징선택(feature selection), 특징 가중치 최적화(feature weighting), 사례선택(instance selection), 모수 최적화(parameter optimization) 등의 방법이 있는데 선행연구에서 가장 많이 사용된 것은 특징선택에 두 기법을 결합하는 방식이다. 본 연구에서도 기계학습 기법을 재무 문제에 활용함에 있어서 최적의 특징선택을 위해 시뮬레이티드 어니일링을 결합하는 방식을 사용한다. 본 연구에서 제안된 기법의 유용성을 확인하기 위하여 실제 재무분야의 데이터를 활용하여 예측 정확도를 확인하였으며 그 결과를 통하여 제안하는 모형의 유용성을 확인할 수 있었다.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.847-854
• /
• 2022

The aim of this study is to develop a Named Entity Recognition (NER) model to automatically identify construction-related organizations from news articles. This study collected news articles using web crawling technique and construction-related organizations were labeled within a total of 1,000 news articles. The Bidirectional Encoder Representations from Transformers (BERT) model was used to recognize clients, constructors, consultants, engineers, and others. As a pilot experiment of this study, the best average F1 score of NER was 0.692. The result of this study is expected to contribute to the establishment of international business strategies by collecting timely information and analyzing it automatically.

• Environmental Engineering Research
• /
• v.21 no.3
• /
• pp.226-232
• /
• 2016

Rainfall is one of crucial factors that impact on our environment. Rainfall data is important in water resources management, flood forecasting, and designing hydraulic structures. However, it is not available in some rural watersheds without rain gauges. Thus, effective ways of interpolating the available records are needed. Despite many widely used spatial interpolation methods, few studies have investigated rainfall center characteristics. Based on the theory that the spatial distribution of convective rainfall event has a definite center with maximum rainfall, we present a mathematical interpolation method to estimate convective rainfall distribution and indicate the rainfall center location and the center rainfall volume. We apply the method to estimate three convective rainfall events in Santa Catalina Island where reliable hydrological data is available. A cross-validation technique is used to evaluate the method. The result shows that the method will suffer from high relative error in two situations: 1) when estimating the minimum rainfall and 2) when estimating an external site. For all other situations, the method's performance is reasonable and acceptable. Since the method is based on a continuous function, it can provide distributed rainfall data for distributed hydrological model sand indicate statistical characteristics of given areas via mathematical calculation.

• Journal of KIISE:Databases
• /
• v.32 no.3
• /
• pp.219-227
• /
• 2005

In this paper, we propose an efficient algorithm which generates frequent itemsets by only one database scanning. A frequent itemset is subset of an itemset which is accessed by a transaction. For each item, if informations about transactions accessing the item are exist, it is possible to generate frequent itemsets only by the extraction of items haying an identical transaction ID. Proposed method in this paper generates the data structure which stores transaction ID for each item by only one database scanning and generates 2-frequent itemsets by using the hash technique at the same time. k(k$\geq$3)-frequent itemsets are simply found by comparing previously generated data structure and transaction ID. Proposed algorithm can efficiently generate frequent itemsets by only one database scanning .