• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1275-1292
• /
• 2021

As a fine-grained classification problem, aspect-level sentiment classification predicts the sentiment polarity for different aspects in context. To address this issue, researchers have widely used attention mechanisms to abstract the relationship between context and aspects. Still, it is difficult to effectively obtain a more profound semantic representation, and the strong correlation between local context features and the aspect-based sentiment is rarely considered. In this paper, a hybrid attention capsule network for aspect-level sentiment classification (ABASCap) was proposed. In this model, the multi-head self-attention was improved, and a context mask mechanism based on adjustable context window was proposed, so as to effectively obtain the internal association between aspects and context. Moreover, the dynamic routing algorithm and activation function in capsule network were optimized to meet the task requirements. Finally, sufficient experiments were conducted on three benchmark datasets in different domains. Compared with other baseline models, ABASCap achieved better classification results, and outperformed the state-of-the-art methods in this task after incorporating pre-training BERT.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2003년도 제2차 연례학술대회 발표논문집
• /
• pp.86-94
• /
• 2003

Electronically available biological literature has been accumulated exponentially in the course of time. So, researches on automatically acquiring knowledge from these tremendous data by text mining technology become more and more prosperous. However, most of the previous researches are technology oriented and are not well focused in practical extraction target, hence result in low performance and inconvenience for the bio-researchers to actually use. In this paper, we propose a more biology oriented target domain specific text mining system, that is, POSTECH bio-text mining system (POSBIOTM), for signal transduction pathway extraction, especially for G protein-coupled receptor (GPCR) pathway. To reflect more domain knowledge, we specify the concrete target for pathway extraction and define the minimal pathway domain ontology. Under this conceptual model, POSBIOTM extracts interactions and entities of pathways from the full biological articles using a machine learning oriented extraction method and visualizes the pathways using JDesigner module provided in the system biology workbench (SBW) [14]

• 센서학회지
• /
• 제31권5호
• /
• pp.279-285
• /
• 2022

Recently, 3D printing technology has gained increased attention in the manufacturing industry because it allows the manufacturing of complex but sophisticated structures as well as moderate production speed. Owing to advantages of 3D printers, such as flexible design, customization, rapid prototyping, and ease of access, can also be advantageous to sensor developments, 3D printing demands have increased in various active device fields, including sensor manufacturing. In particular, 3D printing technology is of significant interest in tactile sensor development where piezoelectric materials are typically embedded to acquire voltage signals from external stimuli. In regard with piezoelectricity, researchers have worked with various piezoelectric materials to achieve high piezoelectric response, but the structural approach is limited because ceramics have been regarded as challenging materials for complex design owing to their limited manufacturing methods. If appropriate piezoelectric materials and approaches to design are used, sensors can be fabricated with the improved piezoelectric response and high sensitivity that cannot be found in common bulk materials. In this study, various 3D printing technologies, material combinations, and applications of various piezoelectric sensors using the 3D printing method are reviewed.

• 한국군사과학기술학회지
• /
• 제19권6호
• /
• pp.721-729
• /
• 2016

CCN(Content-Centric Network) enables users to retrieve content using the content's name. Researchers face critical challenges in terms of mobility. Since the routing information is part of the content name, when the provider moves, it is necessary to update all the routers routing information. However, this requires significant costs. In this paper, we propose PNPCCN(Proactive Neighbor Pushing CCN), considering the popularity and rarity of mobility support, for providers in CCN environments. Via simulation studies, we demonstrate that our solutions are effective in terms of shorter numbers of retransmitted Interest packets, and average download times and higher delivery ratios during mobility.

• Molecular & Cellular Toxicology
• /
• 제1권1호
• /
• pp.46-51
• /
• 2005

Mutagen X (MX) exists in our drinking water as the bi-products of chlorine disinfection. Being one of the most potent mutagen, it attracted much attention from many researchers. MX and its analogs are tested and modeled by quantitative structure activity relationship (QSAR) methods. As a result, factors affecting this class of compounds have been found to be steric and electrostatic effects. We tried to collect all the data available from the literature. The quantitative structure-activity relationship of a set of 29 MX was analyzed using Molecular Field Analysis (MFA) and Receptor Surface Analysis (RSA). The best models gave $q^{2}=0.918,\;r^{2}=0.949$ for MFA and $q^{2}=0.893,\;r^{2}=0.954$ for RSA. The models indicate that an electronegative group at C6 position of the furanone ring increases mutagenicity.

• Management Science and Financial Engineering
• /
• 제11권3호
• /
• pp.47-60
• /
• 2005

Well-known e-CRM strategy is to focus on profitable customers and pay less attention to unprofitable ones. Moreover, some researchers recommend not serving unprofitable ones any more. However, it often neglects customers indirect value. Deselecting unprofitable customers can raise the issue of bad word-of-mouth publicity especially in the age of the Internet. Some studies pointed out that a customers decision to buy a product or service is often strongly influenced by others. In this paper, we consider customers' word-of-mouth effect on quality learning of inexperienced customers. We show that firms implementing e-CRM must take the effect into the consideration when deselecting unprofitable customers.

• Molecules and Cells
• /
• 제46권2호
• /
• pp.120-129
• /
• 2023

Recent technical advances have enabled unbiased transcriptomic and epigenetic analysis of each cell, known as "single-cell analysis". Single-cell analysis has a variety of technical approaches to investigate the state of each cell, including mRNA levels (transcriptome), the immune repertoire (immune repertoire analysis), cell surface proteins (surface proteome analysis), chromatin accessibility (epigenome), and accordance with genome variants (eQTLs; expression quantitative trait loci). As an effective tool for investigating robust immune responses in coronavirus disease 2019 (COVID-19), many researchers performed single-cell analysis to capture the diverse, unbiased immune cell activation and differentiation. Despite challenges elucidating the complicated immune microenvironments of chronic inflammatory diseases using existing experimental methods, it is now possible to capture the simultaneous immune features of different cell types across inflamed tissues using various single-cell tools. In this review, we introduce patient-based and experimental mouse model research utilizing single-cell analyses in the field of chronic inflammatory diseases, as well as multi-organ atlas targeting immune cells.

• 한국콘텐츠학회:학술대회논문집
• /
• 한국콘텐츠학회 2008년도 춘계 종합학술대회 논문집
• /
• pp.861-865
• /
• 2008

최근 몇 년간 국가적으로 R&D 분야에 대한 예산 지원이 활발해지면서 국내 연구자들이 과학기술분야의 주요 학술지에 발표하는 논문의 양도 급격히 증가하고 있다. 하지만 이 연구자들이 발표한 논문을 수록하고 있는 학술지가 국내에서 얼마나 활용되고 있는지에 대한 연구는 이루어지지 않고 있다. 따라서 본 연구에서는 국내에서 연구를 수행하는 연구자들이 발표한 논문이 수록된 주요 학술지에 대한 분석을 통해 한국 연구자들이 주로 발표하는 학술지가 무엇인지를 알아보는데 그 목적이 있다. 이를 위하여 Thomson Reuters 사의 SCI(Science Citation Index) 데이터베이스에 수록된 학술지를 그 분석 대상으로 하였으며, 분석 기간은 2004년부터 2006년도까지 3년간의 데이터를 추출하여 분석하였다.

• 한국진공학회:학술대회논문집
• /
• 한국진공학회 2009년도 제38회 동계학술대회 초록집
• /
• pp.122-122
• /
• 2010

ZnO is a widely investigated material for the blue and ultraviolet solid-state emitters and detectors. It has been promoted due to a wide-band gap semiconductor which has large exciton binding energy of 60 meV, chemical stability and low radiation damage. However, there are many problems to be solved for the growth of p-type ZnO for practical device applications. Many researchers have made an efforts to achieve p-type conductivity using group-V element of N, P, As, and Sb. In this letter, we have studied the optical characteristics of the antimony-doped ZnO (ZnO:Sb) thin films by means of photoluminescence (PL), PL excitation, temperature-dependent PL, and time-resolved PL techniques. We observed donor-to-acceptor-pair transition at about 3.24 eV with its phonon replicas with a periodic spacing of about 72 meV in the PL spectra of antimony-doped ZnO (ZnO:Sb) thin films at 12 K. We also investigate thermal activation energy and carrier recombination lifetime for the samples. Our result reflects that the antimony doping can generate shallow acceptor states, leading to a good p-type conductivity in ZnO.

• International Journal of Knowledge Content Development & Technology
• /
• 제11권3호
• /
• pp.63-79
• /
• 2021

The KJC and KCI services operated by the National Research Foundation of Korea (NRF) need to refer to the global citation index database to set the service direction that researchers want and develop new functions. Therefore, in this study, the usability of Korea's KJC and KCI web databases was evaluated. In addition, FGI verification was performed for a group of experts on KJC and KCI improvement measures derived by academic information experts. The evaluation committee consisted of the researcher group, the society related group, the publisher (academic information distribution organization) group, and the policy-making organization group. A total of 16 experts performed usability evaluation and service improvement proposals for KJC and KCI services. As a result of usability evaluation of KJC and KCI web database using SUS measurement tool, KJC service was evaluated as 73.44, and KCI service was evaluated as good as 64.38. KJC improvement proposals were evaluated positively with 4 points or more for all improvement items, and KCI improvement proposals were positively evaluated with 3.5 points or more for all improvement items. Overall, KJC and KCI error improvements were evaluated with a score of 4 or more, showing strong affirmations. In the future, it is hoped that the improvement proposals and improvements proposed in this study will be applied to the development of KJC and KCI services, so that more quality services will be used by domestic researchers.