• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2021년도 춘계학술대회
• /
• pp.441-443
• /
• 2021

문장 또는 텍스트 유사도란 두 가지 문장의 유사한 정도를 나타내는 척도이다. 텍스트의 유사도를 측정하는 기법으로 자카드 유사도, 코사인 유사도, 유클리디언 유사도, 맨하탄 유사도 등과 같이 있다. 현재 코사인 유사도 기법을 가장 많이 사용하고 있으나 이는 문장에서 단어의 출현 여부와 빈도수에 따른 분석이기 때문에, 의미적 관계에 대한 분석이 부족하다. 이에 우리는 온톨로지를 이용하여 단어 간의 관계를 부여하고, 두 문장에서 공통으로 포함된 단어를 추출할 때 의미적 유사성을 포함함으로써 문장의 유사도에 분석의 효율을 향상하고자 한다.

• 반도체디스플레이기술학회지
• /
• 제20권3호
• /
• pp.113-116
• /
• 2021

As artificial intelligence technologies, including deep learning, develop, these technologies are being introduced to code similarity analysis. In the traditional analysis method of calculating the graph edit distance (GED) after converting the source code into a control flow graph (CFG), there are studies that calculate the GED through a trained graph neural network (GNN) with the converted CFG, Methods for analyzing code similarity through CNN by imaging CFG are also being studied. In this paper, to determine which approach will be effective and efficient in researching code similarity analysis methods using artificial intelligence in the future, code similarity is measured through funcGNN, which measures code similarity using GNN, and Siamese Network, which is an image similarity analysis model. The accuracy was compared and analyzed. As a result of the analysis, the error rate (0.0458) of the Siamese network was bigger than that of the funcGNN (0.0362).

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제7권4호
• /
• pp.295-300
• /
• 2007

The similarity analysis for fuzzy set pair or crisp set pair are carried out. The similarity measure that is based on distance measure is derived and proved. The proposed similarity measure is considered with the help of analysis for uncertainty or certainty part of the membership functions. The usefulness of proposed similarity is verified through the computation of similarity between fuzzy set and crisp set or fuzzy set and fuzzy set. Our results are also compared with those of previous similarity measure which is based on fuzzy number.

• 한국도로학회논문집
• /
• 제18권2호
• /
• pp.51-60
• /
• 2016

PURPOSES : The applicability of the mechanics-based similarity concept (suggested by Feng et al.) for determining scaled variables, including length and load, via laboratory-scale tests and discrete element analysis, was evaluated. METHODS: Several studies on the similarity concept were reviewed. The exact scaling approach, a similarity concept described by Feng, was applied in order to determine an analytical solution of a free-falling ball. This solution can be considered one of the simplest conditions for discrete element analysis. RESULTS : The results revealed that 1) the exact scaling approach can be used to determine the scale of variables in laboratory tests and numerical analysis, 2) applying only a scale factor, via the exact scaling approach, is inadequate for the error-free replacement of small particles by large ones during discrete element analysis, 3) the level of continuity of flowable materials such as SCC and cement mortar seems to be an important criterion for evaluating the applicability of the similarity concept, and 4) additional conditions, such as the kinetics of particle, contact model, and geometry, must be taken into consideration to achieve the maximum radius of replacement particles during discrete element analysis. CONCLUSIONS : The concept of similarity is a convenient tool to evaluate the correspondence of scaled laboratory test or numerical analysis to physical condition. However, to achieve excellent correspondence, additional factors, such as the kinetics of particles, contact model, and geometry, must be taken into consideration.

• 정보과학회 논문지
• /
• 제44권10호
• /
• pp.1019-1025
• /
• 2017

바이너리 유사도 분석은 취약점 분석, 악성코드 분석, 표절 탐지 등에서 사용되고 있는데, 분석대상 함수가 알려진 안전한 함수와 동일하다는 것을 증명해주면 바이너리 코드의 악성행위 분석, 취약점 분석 등의 효율성을 높이는 데에 도움이 될 수 있다. 하지만 기존에는 동일 함수의 서로 다른 버전에 대한 유사도 분석에 대해서 별도로 이루어진 연구가 거의 없었다. 본 논문에서는 바이너리로부터 추출 가능한 함수 정보들을 바탕으로 다양한 방법을 통해 함수 단위의 유사도를 분석하고 적은 시간으로 효율적으로 분석할 수 있는 방안을 모색한다. 특히 OpenSSL 라이브러리의 서로 다른 버전을 대상으로 분석을 수행하여 버전이 다른 경우에도 유사한 함수를 탐지하는 것을 확인한다.

• 복식문화연구
• /
• 제22권5호
• /
• pp.743-759
• /
• 2014

The purpose of this study is to identify the antecedents of consumers' decision postponement on purchasing fast fashion brands. Ongoing search behavior, overchoice confusion, and similarity confusion were considered as antecedents. It was hypothesized that ongoing search behavior influences decision postponement both directly and indirectly through overchoice confusion and similarity confusion. Data were gathered by surveying university students in Seoul, using convenience sampling. Three hundred five questionnaires were used in the statistical analysis, which were exploratory factor analysis using SPSS and confirmatory factor analysis and path analysis using AMOS. Factor analysis proved that ongoing search behavior, overchoice confusion, similarity confusion, and decision postponement were uni-dimensions. Tests of the hypothesized path proved that ongoing search behavior influences decision postponement indirectly through overchoice confusion. In addition, similarity confusion influences decision postponement. The results suggest some confusion reduction strategies for marketers of fast fashion brands. Suggestions for future study are also discussed.

• 한국문헌정보학회지
• /
• 제56권4호
• /
• pp.269-284
• /
• 2022

Exploring research similarity of researchers offers insight on research communities and potential interactions among scholars. While co-authorship is a popular measure for studying research similarity of researchers, it cannot provide insight on authors who have not collaborated yet. In this work, we present novel approach to capture research similarity of authors using citation information. Extensive study is conducted on DATA & KNOWLEDGE ENGINEERING (DKE) publications to demonstrate and compare suggested approach with co-authorship based approach. Analysis result shows that proposed approach distinguishes author relationships that is not shown in co-authorship network.

• Journal of Plant Biotechnology
• /
• 제4권2호
• /
• pp.67-70
• /
• 2002

The RAPD data were used to assess genetic similarity among f grape cultivars. Of the 100 random primers tested on genomic DNA, 10 primers could be selected for Benetic analysis, and the selected primers generated a total of 115 distinct amplification fragments. A similarity matrix was constructed on the basis of the presence or absence of bands. The 7 grape cultivars analyzed with UPGMA were clustered into two groups of A and B. The similarity coefficient value of cultivars was high. The mean similarity index for all pairwise comparisons was 0.851, and ranged from 0.714 ('Rosaki' and 'Black Olympia') to 0.988 ('Kyoho' and 'Daebong'). After due consideration of differences in cultural and morphological characteristics of these two theoretically identical cultivars, it could be deduced that 'Daebong' is a bud sport of 'Kyoho' cultivar.

• Design & Manufacturing
• /
• 제17권3호
• /
• pp.34-40
• /
• 2023

In this study, real-time collection of mold vibration signals during injection molding processes was achieved through IoT devices installed on the mold surface. To analyze changes in the collected vibration signals, injection molding was performed under six different process conditions. Analysis of the mold vibration signals according to process conditions revealed distinct trends and patterns. Based on this result, cosine similarity was applied to compare pattern changes in the mold vibration signals. The similarity in time and acceleration vector space between the collected data was analyzed. The results showed that under identical conditions for all six process settings, the cosine similarity remained around 0.92±0.07. However, when different process conditions were applied, the cosine similarity decreased to the range of 0.47±0.07. Based on these results, a cosine similarity threshold of 0.60~0.70 was established. When applied to the analysis of mold vibration signals, it was possible to determine whether the molding process was stable or whether variations had occurred due to changes in process conditions. This establishes the potential use of cosine similarity based on mold vibration signals in future applications for real-time monitoring of molding process changes and anomaly detection.

• 한국의류산업학회지
• /
• 제15권6호
• /
• pp.1010-1017
• /
• 2013

Current global textiles and fashion industries have gradually shifted focus to high value-added, high sensibility, and multi-functional products based on new human-friendliness and sustainable growth technologies. Textile design CAD systems have been developed in conjunction with computer hardware and software sector advances. This study compares the patterns or images of actual woven fabrics and virtual fabrics prepared with a textile design CAD system. In this study, several weave structures (such as fancy yarn weave and patterns) were prepared with a shuttle loom. The woven textile images were taken using a CCD camera. The same weave structure data and yarn data were fed into a textile design CAD system in order to simulate fabric images as similarly as possible. Similarity Index analysis methods allowed for an analysis of the index between the actual fabric specimen and the simulated image of the corresponding fabric. The results showed that repeated small pattern weaves provide superior similarity index values than those of a fancy yarn weave that indicate some irregularities due to fancy yarn attributes. A Complex Wavelet Structural Similarity(CW-SSIM) index resulted in a better index than other methods such as Multi-Scale(MS) SSIM, and Feature Similarity(FS) SSIM, across fabric specimen images. A correlation analysis of the similarity index based on an image analysis and a similarity evaluation by panel members was also implemented.