• 한국멀티미디어학회논문지
• /
• 제24권12호
• /
• pp.1673-1681
• /
• 2021

Item cold start is a well studied problem in the research field of recommender systems. Still, many existing collaborative filters cannot recommend items accurately when only a few user-item interaction data are available for newly introduced items (Cold items). We propose a interaction feature prediction method to mitigate item cold start problem. The proposed method predicts the interaction features that collaborative filters can calculate for the cold items. For prediction, in addition to content features of the cold-items used by state-of-the-art methods, our method exploits the interaction features of k-nearest content neighbors of the cold-items. An attention network is adopted to extract appropriate information from the interaction features of the neighbors by examining the contents feature similarity between the cold-item and its neighbors. Our evaluation on a real dataset CiteULike shows that the proposed method outperforms state-of-the-art methods 0.027 in Recall@20 metric and 0.023 in NDCG@20 metric.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권12호
• /
• pp.4508-4530
• /
• 2021

Today distributed data storage service are being widely used. However lack of proper means of security makes the user data vulnerable. In this work, we propose a Randomized Block Size (RBS) model for secure data storage in distributed environments. The model work with multifold block sizes encrypted with the Chinese Remainder Theorem-based RSA (C-RSA) technique for end-to-end security of multimedia data. The proposed RBS model has a key generation phase (KGP) for constructing asymmetric keys, and a rand generation phase (RGP) for applying optimal asymmetric encryption padding (OAEP) to the original message. The experimental results obtained with text and image files show that the post encryption file size is not much affected, and data is efficiently encrypted while storing at the distributed storage server (DSS). The parameters such as ciphertext size, encryption time, and throughput have been considered for performance evaluation, whereas statistical analysis like similarity measurement, correlation coefficient, histogram, and entropy analysis uses to check image pixels deviation. The number of pixels change rate (NPCR) and unified averaged changed intensity (UACI) were used to check the strength of the proposed encryption technique. The proposed model is robust with high resilience against eavesdropping, insider attack, and chosen-plaintext attack.

• 공학교육연구
• /
• 제26권3호
• /
• pp.35-41
• /
• 2023

The study was conducted to develop two parallel short forms for the Convergent thinking and Problem solving questionnaires which are part of H University's core competency diagnostic tools, based on Multi-Item Response Theory. Item responses of 2,580 students were analyzed using Graded Response Model(GRM) to determine item difficulty and discrimination of each item. The research results are as follows. Two parrallel short tests were developed for the Convergent thinking questionnaire consisting of 12 items which were originally 17 items. Likewise, the Problem solving questionnaire, which originally consisted of 15 questions, was divided into two parallel short forms, each consisting of 9 items. The reliability of the shortened parallel tests was confirmed through internal consistency analysis, and their similarity to the original tests was established through correlation analysis. This study contributed to quality management of competency-based education and programs at H University by developing shortened tests. Based on the results, implications were presented as well as limitations and discussions.

• 산경연구논집
• /
• 제14권7호
• /
• pp.1-7
• /
• 2023

Purpose: The present research is to explore and investigate the meaningful difference between job-based pay and skill-based pay in the current literature to shed light on the advantages and caveats of two compensation approach, providing direction for human resource practitioners looking to develop efficient and fair compensation plans. Research design, data and methodology: The methodology based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) was used for this research. The PRISMA technique is generally considered the gold standard since it guarantees openness, repeatability, and rigor. Studies comparing salaries based on work duties versus those based on individual skills published in English and accepted for publication in peer-reviewed journals are eligible for inclusion. Results: According to the comprehensive literature analysis, the present research provides clear comparison between Job-based pay and skilled-based pay, pointing out a similarity and four differences (1. Criteria for pay determination, 2. Flexibility, 2. Job Scope, and 4. Career progression. Conclusions: The result of the current research implies that human resource professionals would be well to learn the effects of various compensation structures on career advancement, and suggests for them that employees can take advantage of chances to advance in their careers based on right compensation schemes.

• 한국축산식품학회지
• /
• 제44권2호
• /
• pp.269-283
• /
• 2024

Cultured meat is one of the research areas currently in the spotlight in the agricultural and livestock industry, and refers to cells obtained from livestock that are proliferated and differentiated and processed into edible meat. These cell-cultured meats are mainly studied at the lab-scale by culturing them in flasks, and for commercial use, they are produced using scaffolds that mimic cell supports. Scaffolds are broadly divided into fiber scaffolds, hydrogels, and micro-carrier beads, and these are classified according to processing methods and materials. In particular, a scaffold is essential for mass production, which allows it to have appearance, texture, and flavor characteristics similar to meat. Because cultured meat is cultured in a state where oxygen is blocked, it may be lighter in color or produce less flavor substances than edible meat, but these can be compensated for by adding natural substances to the scaffolds or improving fat adhesion. In addition, it has the advantage of being able to express the texture characteristics of the scaffolds that make up the meat in various ways depending on the materials and manufacturing methods of the scaffolds. As a result, to increase consumers' preference for cultured meat and its similarity to edible meat, it is believed that manufacturing scaffolds taking into account the characteristics of edible meat will serve as an important factor. Therefore, continued research and interest in scaffolds is believed to be necessary.

• 한국의류학회지
• /
• 제48권4호
• /
• pp.632-656
• /
• 2024

This study presents a pattern drafting method for designing a slim-fit blouse for women in their twenties with a D cup bra size and various bust sizes within the standard body type range using the 3D virtual clothing system. Three participants were scanned using a body scanner, and three similar princess blouse patterns were selected. A total of nine virtual blouses were produced, with three sets made for each participant. In the initial evaluations, Pattern C received the highest score. The second pattern was designed by mending the flaws in Pattern C. However, in the eases of the front and back busts, back width and back waist were still insufficient and sleeve length was short. In the similarity evaluation between virtual and real blouses, the overall appearance was similar. Most of the corrected elements received significantly higher ratings in the third real blouse. However, the ease of the rear bust area was slightly large. In general, when drafting blouse patterns for women with large bust sizes, it is essential to increase the ease in the front bust, reduce the ease in the back bust, enlarge the waist dart, and decrease the armhole depth.

• Geomechanics and Engineering
• /
• 제38권5호
• /
• pp.487-495
• /
• 2024

In this study, a series of shaking table model tests were performed to evaluate the dynamic earth pressure acting on pile foundation during liquefaction. The dynamic earth pressure acting on piles were evaluated with depth and pile diameters comparing with excess pore water pressure, it means that the kinematic load effect plays a substantial role in dynamic pile behavior during liquefaction. The dynamic earth pressure acting on pile foundations with mass exhibited significant similarity to those without upper mass. Analyzing the non-fluctuating and fluctuating components of both excess pore water pressure and dynamic earth pressure revealed that the non-fluctuating component has a dominant influence. In case of non-fluctuating component, dynamic earth pressure is larger than excess porewater pressure at same depth, and the difference increased with depth and pile diameter. However, in the case of the fluctuating component, the earth pressure tended to be smaller than the excess pore water pressure as the depth increased. Based on the results of a series of studies, it can be concluded that the dynamic earth pressure acting on the pile foundation during liquefaction is applied up to 1.5 times the excess pore water pressure for the non-fluctuating component and 0.75 times the excess pore water pressure for the fluctuating component.

• Applied Microscopy
• /
• 제50권
• /
• pp.23.1-23.9
• /
• 2020

We propose an effective deep learning model to denoise scanning transmission electron microscopy (STEM) image series, named Noise2Atom, to map images from a source domain 𝓢 to a target domain 𝓒, where 𝓢 is for our noisy experimental dataset, and 𝓒 is for the desired clear atomic images. Noise2Atom uses two external networks to apply additional constraints from the domain knowledge. This model requires no signal prior, no noise model estimation, and no paired training images. The only assumption is that the inputs are acquired with identical experimental configurations. To evaluate the restoration performance of our model, as it is impossible to obtain ground truth for our experimental dataset, we propose consecutive structural similarity (CSS) for image quality assessment, based on the fact that the structures remain much the same as the previous frame(s) within small scan intervals. We demonstrate the superiority of our model by providing evaluation in terms of CSS and visual quality on different experimental datasets.

• 한국육종학회지
• /
• 제42권5호
• /
• pp.525-533
• /
• 2010

본 연구는 사과 품종의 유전적 다양성을 분석하여 육종의 기초 자료로 활용하기 위하여 최근에 국내에서 육성된 품종 및 도입품종을 포함한 34품종을 대상으로 RAPD와 SSR 분석을 수행하였다. RAPD분석에서 총 37종의 선발된 임의 primer를 분석하여 193개의 다형성 밴드(36.2%)를 얻었으며, 평균 다형성 밴드 수는 5.6개였다. SSR 마커 26종을 이용하여 분석한 결과 총 112개의 대립인자가 확인되었고, 마커 당 대립인자 수는 평균 4.3개였다. 유전적 다양성(PIC 값)은 평균 0.843이었고 범위는 0.536(CH03d12)-0.952(CH04c06)였다. RAPD와 SSR분석에서 획득된 305개의 다형성 밴드를 이용하여 UPGMA(비가중 평균결합) 방식으로 유사도 및 집괴분석을 수행한 결과 유전적 유사도 0.640를 기준으로 4개의 그룹으로 분류되었다. 제1그룹에는 '서광'이 단독으로 분류되었고, 제2그룹에는 12품종이 속하였는데 'Spur Earliblaze'와 'Jonathan'을 제외하고 대부분 'Golden delicious'를 교배친으로 이용하여 육성된 품종이 분포하는 것으로 나타났다. 제3그룹에는 'Fuji'와 'Fuji'를 교배친으로 이용되여 선발된 품종 및 그의 아조변이 품종 등 13개 품종이 속하였고 제4그룹에는 '홍로', '감홍', '새나라' 등 8품종이 포함되었다. 품종간 유전적 유사도는 0.529-0.987의 범위로 평균 유전적 유사도는 0.647이었다. 가장 높은 유사도 값(0.987)을 나타낸 품종은 '화랑'과 '단홍' 품종 간이었고 가장 낮은 유사도 지수(0.529)를 나타낸 품종은 '서광'과 '화랑' 품종 간이었다. 본 실험을 통해 사과 34품종 간의 유연관계는 알려진 pedigree와 일치하는 것을 알 수 있었다.

• 한국방사선학회논문지
• /
• 제13권1호
• /
• pp.111-118
• /
• 2019

흉부 디지털 단층영상합성장치는 기존 DR의 낮은 깊이 해상도, CT의 높은 피폭선량 문제를 해결할 수 있는 획기적인 영상장치로 대두되고 있다. 그러나 제한된 스캔 각도로 인해 프로젝션이 X 선 소스 동작 방향으로 흉부를 완전히 포함 할 수 없어 재구성 된 슬라이스의 위, 아래 방향 경계를 따라 강도의 불연속성이 발생하게 되는데 이러한 현상을 잘림 아티팩트 (Truncation artifact)라고 한다. 이 연구의 목적은 가중 정규화 접근법을 사용하여 잘림 아티팩트를 줄이고 리스템에서 개발한 프로토 타입 흉부 디지털 단층영상합성장치 시스템에 대한 이 접근법의 성능을 평가하는 것이다. 이 시스템의 source-to-image distance는 1100 mm 이고 X 선원의 회전 중심은 검출기 표면에서 100mm 위로 설정되었다. LUNGMAN 팬텀을 사용하여 ${\pm}20^{\circ}$의 투영 뷰를 $1^{\circ}$ 간격으로 41장을 얻은 후, filtered back projection 알고리즘으로 재구성했다. 정량적 평가를 위하여 시뮬레이션을 이용하여 기준영상을 재구성 후 peak signal to noise ratio와 structure similarity index 값을 평가하였으며 실제 실험 데이터를 이용하여 mean value of specific direction 값을 평가하였다. 시뮬레이션 결과로 아티팩트 보정 전 일반적인 filtered back projection 알고리즘으로 재구성 한 영상과 비교하여 peak signal to noise ratio값과 structure similarity index값 모두 각각 증가하였으며, 실제 실험 재구성 영상의 mean value of specific direction 결과는 아티팩트의 영향이 감소됨을 확인할 수 있었다. 결론적으로, 가중 정규화 방법은 잘림 아티팩트를 줄임으로써 진단의 어려움을 발생시키는 가능성을 개선시킬 수 있는 방법으로 사료된다.