1,805 research outputs found
The regional promotion through Moe about the potentialities and its limitations of Moe-Okoshi
No full textIn this paper, the athor tries to discuss Moe-Okoshi, a kind of regional promotion through Moe.
Moe is an emotion of attachment to fictitious objects, like characters in animations, comics, or computer games. Originally, Moe was a slang of the Japanese OTAKU culture, but is now globally known as a term for a more common affection for sweeties. In fact, the market of Moe is becoming larger and larger, and attracting much attention of those who have had no connection with the pop culture before. The regional promotion, the main subject of this paper, is also having an intimate relation with Moe. Many novel promotion plans, in which practitioners try to draw the public notice by means of the fascination or curiosity of Moe, have already appeared. Such plans are commonly called Moe-Okoshi.
However, a lot of practices which we have bundled with the name of Moe-Okoshi can be classified into some models in spite of the superficial similarities; the frequent use of pretty animation-like pictures and so on. And each model has a different feature and method. A few of them have already been highlighted as individual examples, but they have been given no opportunity to be discussed systematically in the context of regional promotion.
These circumstances duly considered, it may be very meaningful to analyze the strategies which a practitioner can take toward each Moe-Okoshi model. For this purpose, the author investigates the potentialities and limitations of Moe-Okoshi.論文(Article)departmental bulletin pape
Recovering Reed-Solomon codes privately
No full textWe investigate the problems of privately repairing erasures and evaluating their linear combinations for Reed-Solomon codes with low communication bandwidths. We propose two approaches: one based on hiding subspaces used to form parity-check equations, and another based on multiplying parity-check equations with random polynomials. We also derive a lower bound on the repair bandwidth for the single erasure case under reasonable assumptions about the schemes being used and demonstrate the optimality of the proposed schemes for codes of specific lengths.Ministry of Education (MOE)National Research Foundation (NRF)Submitted/Accepted versionThis research is supported by the National Research Foundation, Singapore under its Strategic Capability Research Centres Funding Initiative, Ministry of Education, Singapore, under its MOE Academic Research Fund Tier 2 Grants MOE-T2EP20121-0007 and MOE-000623-00, Tier 1 Grant RG19/23, Australia Research Council (ARC) DECRA Grant DE180100768, Israel Science Foundation (ISF) under Grant 2462/24. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore
Schoolgirls With Guns The Evolution of Military Moe as an Otaku Subgenre
No full text45 p.The author examines a fusion subgenre of Japanese anime and manga otaku: military moe. Although it was only solidified as a concept in the mid-2000s, military moe anime and manga have exploded in popularity over a relatively short period of time. This paper aims to give a comprehensive history of military moe's creation, as well as brief analysis of important pieces of media. The author considers why military moe, which bears little surface resemblance to traditional military otaku media, was able to find success among a subset of military otaku
SGAT: simplicial Graph attention network
No full textHeterogeneous graphs have multiple node and edge types and are semantically richer than homogeneous graphs. To learn such complex semantics, many graph neural network approaches for heterogeneous graphs use metapaths to capture multi-hop interactions between nodes. Typically, features from non-target nodes are not incorporated into the learning procedure. However, there can be nonlinear, high-order interactions involving multiple nodes or edges. In this paper, we present Simplicial Graph Attention Network (SGAT), a simplicial complex approach to represent such high-order interactions by placing features from non-target nodes on the simplices. We then use attention mechanisms and upper adjacencies to generate representations. We empirically demonstrate the efficacy of our approach with node classification tasks on heterogeneous graph datasets and further show SGAT's ability in extracting structural information by employing random node features. Numerical experiments indicate that SGAT performs better than other current state-of-the-art heterogeneous graph learning methods.Economic Development Board (EDB)Ministry of Education (MOE)Nanyang Technological UniversitySubmitted/Accepted versionThe frst author is supported by Shopee Singapore Private Limited under the Economic Development Board Industrial Postgraduate Programme (EDB IPP). The programme is a collaboration between Shopee and Nanyang Technological University, Singapore. The last two authors are supported in part by the Singapore Ministry of Education Academic Research Fund Tier 2 grant MOE-T2EP20220-0002
BrainOOD: out-of-distribution generalizable brain network analysis
No full textIn neuroscience, identifying distinct patterns linked to neurological disorders, such as Alzheimer’s and Autism, is critical for early diagnosis and effective intervention. Graph Neural Networks (GNNs) have shown promising in analyzing brain networks, but there are two major challenges in using GNNs: (1) distribution shifts in multi-site brain network data, leading to poor Out-of-Distribution (OOD) generalization, and (2) limited interpretability in identifying key brain regions critical to neurological disorders. Existing graph OOD methods, while effective in other domains, struggle with the unique characteristics of brain networks. To bridge these gaps, we introduce BrainOOD, a novel framework tailored for brain networks that enhances GNNs’ OOD generalization and interpretability. BrainOOD framework consists of a feature selector and a structure extractor, which incorporates various auxiliary losses including an improved Graph Information Bottleneck (GIB) objective to recover causal
subgraphs. By aligning structure selection across brain networks and filtering noisy features, BrainOOD offers reliable interpretations of critical brain regions. Our approach outperforms 16 existing methods and improves generalization to OOD subjects by up to 8.5%. Case studies highlight the scientific validity of the patterns extracted, which aligns with the findings in known neuroscience literature. We also propose the first OOD brain network benchmark, which provides a foundation for future research in this field. Our code is available at
https://github.com/AngusMonroe/BrainOOD.Ministry of Education (MOE)Published versionThis research/project is supported by the Ministry of Education, Singapore under its MOE Academic Research Fund Tier 2 (STEM RIE2025 Award MOE-T2EP20220-0006) and Tier 1 (RG16/24). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of the Ministry of Education, Singapore. YQ and JC are supported by Research Grants 8601116, 8601594, and 8601625 from the UGC of Hong Kong
Genetic algorithm LQG and neural network controllers for gust response alleviation of flying wing unmanned aerial vehicles
No full textIn this paper, a genetic algorithm linear quadratic Gaussian controller (GA-LQG) and an artificial neural network (ANN) controller are implemented for gust response alleviation of lightweight flying wings undergoing body-freedom oscillations. A state–space aeroelastic model has been formulated by coupling the unsteady vortex lattice method for aerodynamics with finite-element based structural dynamics. The model is subsequently reduced using balanced truncation to improve computational efficiency during controller synthesis. Open-loop simulations show that the flying wing experiences large changes in pitching angles during gusts. For GA-LQG controller, the LQG weights are optimised using a genetic algorithm, maximising a defined fitness function. Generally, the GA-LQG controller reduces the plunge displacements by up to 94.2% while damping out wingtip displacements for discrete and continuous gusts. Similarly, the ANN controller effectively regulates both the plunge displacements and wingtip displacements, including gust cases that are not presented during the ANN training phase. The ANN controller is more effective in correcting wingtip displacements during discrete gusts than the GA-LQG controller, while the opposite is true for the continuous gust cases. The ANN controller offers several advantages over the GA-LQG controller, including the elimination of the need for a Kalman filter for full state estimation and offers a non-linear control solution.Agency for Science, Technology and Research (A*STAR)Ministry of Education (MOE)Nanyang Technological UniversitySubmitted/Accepted versionThe first author acknowledges the support from Nanyang Technological University, Singapore for providing the Nanyang President’s Graduate Scholarship. The second author acknowledges the support by A*STAR under its MTC IAF-PP Programme (M23L5a0002) and MOE Tier 1 (RG142/23) and MOE Tier 2 (MOE-T2EP50123-0003)
[[alternative]]Exploring the rationale, contents and models of Environmental Education projects sponsored by MOE in 2002 and 2003
No full text[[abstract]]The environmental education projects sponsored by Ministry of Education (MOE) were the subject of content analysis in this study. The concepts used for analyzing the contents of the EE projects were EE rationale (environmental conservation, safe and healthy community, social justice and intergenerational justice, and environmental ethics), EE principles (totality, lifelong learning, interdisciplinary, participating, issue oriented, holistic, and partnership), EE goals (awareness, knowledge, attitude, skills, and action), and EE activities. The purpose of this study was to explore the rationale, contents and models of EE projects sponsored by MOE.
The result found that the contents and target audiences of the EE projects sponsored by MOE were more diversified. The participants of the 83 projects were from primary school students to general public, and the subjects of the EE projects included eco-tourism, bio-diversity, environmental conservation, extinction, environmental interpretation, green building, Agenda 21, sustainable development, green life, green spirit, ocean conservation, wet land, and aboriginal culture.
In addition, the strategies of these projects to implement the EE projects could be classified as following: symposium, teacher training workshop, propaganda, ecological experiencing activities, green life activities, and mass media.
Dowker complex based machine learning (DCML) models for protein-ligand binding affinity prediction
No full textWith the great advancements in experimental data, computational power and learning algorithms, artificial intelligence (AI) based drug design has begun to gain momentum recently. AI-based drug design has great promise to revolutionize pharmaceutical industries by significantly reducing the time and cost in drug discovery processes. However, a major issue remains for all AI-based learning model that is efficient molecular representations. Here we propose Dowker complex (DC) based molecular interaction representations and Riemann Zeta function based molecular featurization, for the first time. Molecular interactions between proteins and ligands (or others) are modeled as Dowker complexes. A multiscale representation is generated by using a filtration process, during which a series of DCs are generated at different scales. Combinatorial (Hodge) Laplacian matrices are constructed from these DCs, and the Riemann zeta functions from their spectral information can be used as molecular descriptors. To validate our models, we consider protein-ligand binding affinity prediction. Our DC-based machine learning (DCML) models, in particular, DC-based gradient boosting tree (DC-GBT), are tested on three most-commonly used datasets, i.e., including PDBbind-2007, PDBbind-2013 and PDBbind-2016, and extensively compared with other existing state-of-the-art models. It has been found that our DC-based descriptors can achieve the state-of-the-art results and have better performance than all machine learning models with traditional molecular descriptors. Our Dowker complex based machine learning models can be used in other tasks in AI-based drug design and molecular data analysis.Ministry of Education (MOE)Nanyang Technological UniversityPublished versionThis work was supported in part by Nanyang Technological University Startup Grant M4081842 and Singapore Ministry of Education Academic Research fund Tier 1 RG109/19, MOET2EP20120- 0013 and MOE-T2EP20220-0010. The first author (XL) was supported by Nankai Zhide foundation. The second author (HF) was supported by Natural Science Foundation of China (NSFC grant no. 11931007, 11221091, 11271062, 11571184). The third author (JW) was supported by Natural Science Foundation of China (NSFC grant no. 11971144) and High-level Scientific Research Foundation of Hebei Province
Data watermarking for sequential recommender systems
No full textIn the era of large foundation models, data has become a crucial component in building high-performance AI systems. As the demand for high-quality and large-scale data continues to rise, data copyright protection is attracting increasing attention. In this work, we explore the problem of data watermarking for sequential recommender systems, where a watermark is embedded into the target dataset and can be detected in models trained on that dataset. We focus on two settings: dataset watermarking, which protects the ownership of the entire dataset, and user watermarking, which safeguards the data of individual users. We present a method named Dataset Watermarking for Recommender Systems (DWRS) to address them. We define the watermark as a sequence of consecutive items inserted into normal users' interaction sequences. We define a Receptive Field (RF) to guide the inserting process to facilitate the memorization of the watermark. Extensive experiments on five representative sequential recommendation models and three benchmark datasets demonstrate the effectiveness of DWRS in protecting data copyright while preserving model utility.Ministry of Education (MOE)Published versionThis research is supported by the Ministry of Education, Singapore, under its Academic Research Fund (Tier 2 Award MOE-T2EP20221-0013 and Tier 1 Award (RG20/24)). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of the Ministry of Education, Singapore. The Australian Research Council partially supports this work under the streams of Future Fellowship (Grant No. FT210100624), the Discovery Project (Grant No. DP240101108), and the Linkage Projects (Grant No. LP230200892 and LP240200546)
Structure-based multilevel descriptors for high-throughput screening of elastomers
No full textTo discover new materials, high-throughput screening (HTS) with machine learning (ML) requires universally available descriptors that can accurately predict the desired properties. For elastomers, experimental and simulation data in current descriptors may not be available for all candidates of interest, hindering elastomer discovery through HTS. To address this challenge, we introduce structure-based multilevel (SM) descriptors of elastomers derived solely from molecular structure that is universally available. Our SM descriptors are hierarchically organized to capture both local soft and hard segment structures as well as the global structures of elastomers. With the SM-Morgan Fingerprint (SM-MF) descriptor, one of our SM descriptors, a machine learning model accurately predicts elastomer toughness with a remarkable accuracy of 0.91. Furthermore, an HTS pipeline is established to swiftly screen elastomers with targeted toughness. We also demonstrate the generality and applicability of SM descriptors by using them to construct HTS pipelines for screening elastomers with a targeted critical strain or Young's modulus. The user-friendliness and low computational cost of SM descriptors make them a promising tool to significantly enhance HTS in the search for novel materials.Ministry of Education (MOE)National Research Foundation (NRF)This research/project is supported by the National Research Foundation, Singapore (NRF) under NRF’s Medium Sized Centre: Singapore Hybrid-Integrated Next-Generation μ-Electronics (SHINE) Centre funding programme. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of the National Research Foundation, Singapore. We also acknowledge support from the Ministry of Education (MOE) of Singapore under Academic Research Fund Tier 2 (MOE-T2EP20221-0003)
- …
