Representation Scheme Research Articles

Data-Driven Fuzzy Target-Side Representation for Intelligent Translation System

The encoder–decoder framework has been widely used in various practical artificial intelligence cyber-physical systems, including intelligent translation systems. The decoding process in such a framework usually demands the target-side representation, which is often learned by an autoaggressive decoder to simulate the target context information at the current time-step. However, the autoaggressive decoder only captures the previously generated partial target fragment and fails in simulating the global contextual information. In this article, we propose a new data-driven fuzzy context representation strategy to simulate the global target information. Specifically, we design two fuzzy methods to the global target contextual information, which are bag-of-words of target language generated via a softmax layer from the source-side representation and whole target sentence retrieved from the translation memory according to the source-side representation. Both methods facilitate the autoaggressive decoder to handle the global target context at the current time-step, thereby learning a more effective context vector for the generation of target translation. Extensive experiments on two machine translation tasks demonstrated that the proposed method achieved 3% improvement of BLEU score over a strong baseline.

Effect of Representation Strategies on Secondary School Students’ Mathematics Achievement: A Case Unguja Island, Zanzibar

This study sought to establish the effect of representations strategies on Secondary School students’ Mathematics Achievement in Unguja Island, Zanzibar. Four out of sixteen public secondary schools located in the urban district of Unguja Island were randomly selected. Two of these schools were assigned as the experimental group while the other two were the control group. Participants were students from level three who were sampled based on the type of group to which the school is assigned. Achievement tests (pre-test and post-test) based on mathematics topics of relation and function were used as an instrument for data collection. Data was analyzed quantitatively in terms of descriptive statistics and t-tests. The study established no initial difference in performance between the control and the experimental groups. After the intervention, the experimental group which used multiple representation strategies outperformed the control group. Therefore, the study recommends that teachers should implement multiple representation strategies during the mathematics teaching. Additionally, teachers should be provided with in-service training on proper utilization of varieties of representation strategies in teaching and learning of mathematics for better results to be realized.

A Practical Light Field Representation and Coding Scheme with an Emphasis on Refocusing

Light field images have drawn a great deal of interest owing to their flexible post-processing and versatile functionalities. However, the tremendous data volume has imposed a practical limit in applications, which requires a better way to represent and compress the data. In this paper, we propose a practical representation and coding scheme for light field data with a special emphasis on retaining the refocusing functionality as much as possible. Under the proposed method, light field data are represented and compressed in the form of a rendered all-in-focus image and a depth map. Before encoding, the all-in-focus image is rendered accurately based on the depth map using the proposed difference focus measure. At the decoder, a focal stack comprising multiple images having different focus levels can be reconstructed by applying a defocusing function to the compressed all-in-focus image where the defocusing function parameter depends on the selected depth level of the compressed depth map. Compared with well-known state-of-the-art methods that compress the Fourier Disparity Layer or the focal stack, the proposed coding scheme shows much smaller loss in the refocusing capability (16.2% and 17.8% less than two state-of-the-art methods) and provides PSNR improvements of 1.60㏈ to 2.38㏈ at the same compression ratio.

Comics into Adversary: A consideration of how comics thinking can inform the representational challenges of post-crisis creative writing

What can formalist comics studies contribute to creative writing practice? This consideration of an experience in unconsciously applying comics thinking to creative writing shows how notions familiar to comics studies can enrich creative writing. The essay articulates the challenges of writing contemporary gay male “post-crisis” fiction, which troubles the foundations of many representational strategies familiar to creative writers, raising questions about the relationship between what can be shown and what can be known. In comics, where spatialised relationships are foregrounded and help guide representational strategies such as focalisation and description, these foundations become decentred and malleable. Yet, rather than using comics thinking to resolve problems in creative writing – the temptation of applied practice – this paper shows how looking at representational strategies across media can allow challenges to constitute a piece of creative writing and therefore stop being problems to be “solved” but rather to be negotiated within a particular work. The discussion contributes to comics studies and creative writing through highlighting echoes and distortions between the two linked disciplines; theory and method from one creative discipline may be formally applied to another, but the benefits of using cognate disciplines to “think through” problems can also be indirect and discursive.

Генеалогия прежде всего! Формирование сословной группы клириков в раннесредневековом ирландском обществе

This article analyzes genealogies as a strategy of representation of clergy in Early Medieval Ireland. The Irish genealogies, especially the genealogies of saints were, as D. Ó Corráin put it, a product of “dynastically-minded clergy”: this hereditary clergy arose from the 7th century on, at latest. Trying to represent the founder saint in the most favourable light and to confirm their own privileges and rights to church property, the compilers of genealogies often used the following devices: (1) suppressing the saint’s foreign origin (St Ailbe and others); (2) portraying the saint as a member of the local ruler’s family or as a relative of an incumbent abbot or an abbatial dynasty (St Rúadán of Lothra); (3) constructing the saint’s maternal genealogy, thereby forging the necessary dynastic connections. All the three strategies were applied in the Irish legend of Saint Gregory the Great: Gregory is presented as a son of an Irishman, who was allegedly a relative of the previous Pope, and his mother was said to be a Pope’s daughter.

Label consistency-based deep semisupervised NMF for tumor recognition

Tumor has become a hot topic in the field of image processing and pattern recognition. Gene expression data is an important way to study tumor. Since gene expression data is characterized by high-dimensional small samples, it is one of the key steps to extract discriminative gene features to distinguish different tumor types. Nonnegative matrix factorization (NMF) is an unsupervised feature representation method that does not depend on the label information of data. NMF can achieve nonlinear dimension reduction and is widely used in tumor recognition. Considering that some data may carry label information, models’ feature representation capability will be improved if the label information can be used effectively. Therefore, this paper proposes a semisupervised NMF model based on label consistency (LC-NMF), which uses both labeled and unlabeled data to obtain feature representations. Furthermore, to alleviate the sensitivity of the NMF model to initial values and excavate the deep features of data, a label consistency-based deep semisupervised NMF model (LC-DNMF) is constructed, which combines the LC-NMF model with the layer-by-layer pretraining and multilayer representation strategy in deep learning. The performance of the proposed models (i.e., LC-NMF and LC-DNMF) is verified by applying them to the tumor recognition tasks. The experimental results on seven datasets show that the two models achieve good results and can obtain competitive recognition accuracies compared with the state-of-the-art methods. Furthermore, the performance of the LC-DNMF model outperforms that of the LC-NMF model, which verifies the effectiveness of introducing the layer-by-layer pretraining and multilayer representation strategy.

Neuromimesis: Picturing the Humanities Picturing the Brain.

What do neuroscientific visualizations of mental functioning depict? This article argues that neuroscientific imaging from Santiago Ramón y Cajal’s pen and ink drawings onward falls within the mimetic tradition, that dealing with the artistic representation of reality. Cajal’s iconic images of pyramidal neurons and glial cells surprisingly suggest a non-realist approach to picturing the brain and the mind that opens a new methodological link between humanities and neurosciences. In it, aesthetic works offer a perspective on mimetic practices in neurosciences, providing insight into representational strategies that make otherwise invisible psychic phenomena observable. This approach draws needed attention to the role of metaphor in neuroscientific research. It also reimagines how interdisciplinary scholarship might engage with works of art. While it is a common practice to read humanities objects featuring the brain and/or the mind in terms of their neuroscientific content, films like The Headless Woman (La mujer sin cabeza, dir. Martel, 2008), explored here, show that doing so can easily inhibit interpretations with greater explanatory bearing. Together, Cajal’s images and Martel’s film help elaborate a fresh methodological paradigm—distinct from that of neuropsychoanalysis—that situates aesthetic objects as a long-neglected tool for studying the brain by virtue of (not despite) their imaginative investments.

Novel material representation method via a deep learning model for multi-scale topology optimization

In this paper, a novel deep learning-aided material representation scheme for multi-scale topology optimization is proposed. This method shows that it is possible to determine a general-purpose mapping from a low-dimensional variable to the image of microstructures. A deep generative model learns features from microstructural images to find manifolds defined in the low-dimensional latent space, then a regression model is trained to fit the equivalent material properties. After training, the generator and predictor networks are integrated into the multi-scale topology optimization process to reduce the number of design variables and replace the homogenization computation, respectively. With the proposed material representation method, the optimization algorithm converges faster, while automatically satisfying complicated geometrical restrictions without any additional constraints. Due to the generator network, the microstructures can be interpolated over the latent manifold. It enables the multi-scale topology optimization can be conducted over an irregular design domain with unstructured mesh. The effectiveness of this method is tested with two simple manually designed microstructures and a complex one obtained by inverse homogenization, and its performance is discussed based on the number of design variables, computational efficiency, and optimized multi-scale design results. The optimization performance tends to be improved as the latent dimensions increase. The results show that, on average, the elapsed time per iteration of the proposed method is close to two percent of that of conventional methods. By means of the ability to interpolate microstructures, a high-resolution full-scale realization can be obtained from a lower-resolution design, and the proposed method can utilize an unstructured mesh in multi-scale topology optimization.

How Not to Forget

Can interior design facilitate remembrance? While the aesthetic and literary genre of speculative fiction projects contemporary political, social, and econometric structures into dystopian futures, speculative history turns this fictional and factual focus to the past. It uses photographs, textual records, oral testimony, and interpretation and extrapolation to intervene in existing narratives. This process involves engaging the absences and gaps in the ephemera housed in institutional archives that work to construct official records of historic events. This text-based essay explores the potential of engaging speculative history as a form of interior design praxis. How can lost narratives, hidden voices, architectural violence, and other concealed historical traumas interwoven with the built environment be resurfaced through strategies of interior visualisation? To address this prompt, four projects from the studio course Such a Place as Memory are taken as case studies to explore how the interior can be situated between public and personal forms of memory. By intervening in historical records and extrapolating familial histories, each work reveals the potential of digital or analogue tools to counter absences in the archive. Stories of communities at themargins or subject to violent political efforts are captured through representational strategies of advanced digital software, orthographic drawing, and architectural modelling. In each, speculation, interpretation, and intervention are centred as keyelements informed by theorist Marianne Hirsch’s concept of postmemory. In my reading, the speculative interior and its strategies of representation may function as an apparatus of memory that can capture and relay unseen or deliberately concealed historical narratives into the future.

Cost-based Optimization of Multistore Query Plans

Multistores are data management systems that enable query processing across different and heterogeneous databases; besides the distribution of data, complexity factors like schema heterogeneity and data replication must be resolved through integration and data fusion activities. Our multistore solution relies on a dataspace to provide the user with an integrated view of the available data and enables the formulation and execution of GPSJ queries. In this paper, we propose a technique to optimize the execution of GPSJ queries by formulating and evaluating different execution plans on the multistore. In particular, we outline different strategies to carry out joins and data fusion by relying on different schema representations; then, a self-learning black-box cost model is used to estimate execution times and select the most efficient plan. The experiments assess the effectiveness of the cost model in choosing the best execution plan for the given queries and exploit multiple multistore benchmarks to investigate the factors that influence the performance of different plans.

A parallel double scrambling encryption scheme for MQIR image based on random combination

With the rapid expansion of quantum information and quantum networks, quantum image security research has gradually become a research hotspot in the field of information security. However, in the field of quantum image processing, most encryption algorithms are independent of color components or position space, and the encryption effect is not satisfactory. Therefore, we propose a parallel double scrambling encryption scheme for a multimode quantum image representation (MQIR) image based on random combination, which uses the principle of random combination to encrypt the position and pixel value of the image at the same time, so as to realize the parallel double scrambling of the pixel value and position of the MQIR image. The algorithm is based on specific keys when encrypting, so that the generation of the keys are related to the seeds. At the same time, a certain number of bits are selected on the qubits representing the three-dimensional positions, and NOT gates are added. This is so that the generated pseudo-random sequence is also related to the number of qubits chosen for the three-dimensional position of the quantum image, making it more difficult to decrypt the encrypted image. We give the quantum circuits with parallel double scrambling and inverse scrambling for MQIR quantum images. In the circuit, we divide the image into four sub-blocks and locate them respectively at the end of each sub-block. The encryption key for different sub-blocks depends on the selected seed. This circuit only needs to encrypt four times without iterating. The algorithm is simple and has a large key space. Finally, we perform the correlation coefficient (CC) and the number of pixels change rate (NPCR) analysis between the plaintext image and the encrypted image. The experimental results show that the CC exhibits low correlation between the plaintext image and the encrypted image, and the NPCR between them is extremely high, which verifies the effectiveness and security of the proposed quantum image encryption scheme.

NcRNAInter: a novel strategy based on graph neural network to discover interactions between lncRNA and miRNA.

In recent years, many studies have illustrated the significant role that non-coding RNA (ncRNA) plays in biological activities, in which lncRNA, miRNA and especially their interactions have been proved to affect many biological processes. Some in silico methods have been proposed and applied to identify novel lncRNA-miRNA interactions (LMIs), but there are still imperfections in their RNA representation and information extraction approaches, which imply there is still room for further improving their performances. Meanwhile, only a few of them are accessible at present, which limits their practical applications. The construction of a new tool for LMI prediction is thus imperative for the better understanding of their relevant biological mechanisms. This study proposed a novel method, ncRNAInter, for LMI prediction. A comprehensive strategy for RNA representation and an optimized deep learning algorithm of graph neural network were utilized in this study. ncRNAInter was robust and showed better performance of 26.7% higher Matthews correlation coefficient than existing reputable methods for human LMI prediction. In addition, ncRNAInter proved its universal applicability in dealing with LMIs from various species and successfully identified novel LMIs associated with various diseases, which further verified its effectiveness and usability. All source code and datasets are freely available at https://github.com/idrblab/ncRNAInter.

Evolutionary Search for Complete Neural Network Architectures With Partial Weight Sharing

Neural architecture search (NAS) provides an automatic solution in designing network architectures. Unfortunately, the direct search for complete task-dependent network architectures is laborious since training and evaluating complete neural architectures over a large search space are computationally prohibitive. Recently, one-shot NAS (OSNAS) has attracted great attention in the NAS community because it significantly speeds up the candidate architecture evaluation procedure through weight sharing. However, the full weight sharing training paradigm in OSNAS may result in strong interference across candidate architectures and mislead the architecture search. To alleviate the problem, we propose a partial weight sharing OSNAS framework that directly evolves complete neural network architectures. In particular, we suggest a novel node representation scheme that randomly activates a subset of nodes of the one-shot model in each generation to reduce the weight coupling in the one-shot model. During the evolutionary search, a tailored crossover operator randomly samples the nodes from two parent individuals or a single parent to construct new candidate architectures, thus effectively constraining the degree of weight sharing. Furthermore, we introduce a new mutation operator that replaces the chosen nodes of the one-shot model with randomly generated nodes to enhance the exploratory capability. Finally, we encode a set of pyramidal convolution operations in the search space, enabling the evolved neural networks to capture different levels of details in the images. The proposed method is examined and compared with 26 state-of-the-art algorithms on ten image classification tasks, including CIFAR series, CINIC10, ImageNet, and MedMNIST series. The experimental results demonstrate that the proposed method can computationally much more efficiently find neural architectures that achieve comparable classification accuracy to the state-of-the-art designs.

Non-Local Representation Based Mutual Affine-Transfer Network for Photorealistic Stylization.

Photorealistic stylization aims to transfer the style of a reference photo onto a content photo in a natural fashion, such that the stylized image looks like a real photo taken by a camera. State-of-the-art methods stylize the image locally within each matched semantic region and are prone to global color inconsistency across semantic objects/parts, making the stylized image less photorealistic. To tackle the challenging issues, we propose a non-local representation scheme, constrained with a mutual affine-transfer network (NL-MAT). Through a dictionary-based decomposition, NL-MAT is able to successfully decouple matched non-local representations and color information of the image pair, such that the context correspondence between the image pair is incorporated naturally, which largely facilitates local style transfer in a global-consistent fashion. To the best of our knowledge, this is the first attempt to address the photorealistic stylization problem with a non-local representation scheme, such that no additional models or steps for semantic matching are required during stylization. Experimental results demonstrate that, the proposed method is able to generate photorealistic results with local style transfer while preserving both the spatial structure and global color consistency of the content image.

Rendering Black Queer Operations: Technologies for Knowing, Being, and Living in Dirty Computer

Abstract In 2018, musical artist Janelle Monáe released Dirty Computer, a sci-fi film/visual album accompanying her musical album of the same name. Dirty Computer plays with elements of the visual album genre by embedding conventional music videos within a larger sci-fi narrative, the videos functioning as representations of Monáe's character's supposed “memories.” This essay analyzes representational strategies at play in Dirty Computer, drawing from Black feminist and queer theory, Black cultural studies, and media studies. I argue that together the narrative universe and structure of Dirty Computer strategically depict struggles over living as Black, queer, and/or woman; maintaining sovereignty over othered bodies and selves; and creating Black worlds within and despite an anti-Black world. Thematically, the film offers futuristic visions of queer/women's collectivity and Black queer intimacy. The film's content and structure render an example of Black Queer OS (Operating System); “Queer OS,” as Kara Keeling has described, “seeks to make queer into the logic of ‘an operating system of a larger order’ that unsettles the common senses that secure those presently hegemonic social relations that can be characterized by domination, exploitation, oppression, and other violences.” Dirty Computer reveals surveillance technologies and oppressive institutions as unreliable narrators, incapable of fully or accurately depicting Black queer life. Instead, it is our own bodies and minds that enable technologies for truly knowing and relating as Black, queer, and/or woman. The film demonstrates strategies of subversive social, cultural, and political operation, presenting Black queer subjectivity as a technology of liberated being. Ultimately, Dirty Computer instructs us to operationalize Black queer life, joyfully together, on a larger order.

건강기능식품 종합비타민 패키지 디자인의 기호학적 연구

Due to the Corona 19 (COVID-19) pandemic, consumer health concerns and consumption are leading to a trend of popularization. As a result, brands are developing various marketing strategies to meet consumer needs and gain consumer recognition. Visual expression through package design is a powerful and close communication tool for brands to communicate with consumers. The purpose of this study is to consider the strategy of visual symbolic representation of package design, which can enhance the effect of positive brand image building and association through grasping expression elements and symbolic interpretation and analysis of the meaning transmission of each element. The visual expression element analysis and color analysis through the visual symbols expressed in the package design of the top 10 multivitamin brands in Korea, the United States, and China, and the meaning transmission process of the visual symbols that contain the meaning and symbol. As a result of the research, first of all, a composite compositional design strategy of linguistic and non-verbal (visual) elements of illustrations that can intuitively associate information in the visual expression elements of package design is necessary. is. Secondly, visual expression through color requires a color scheme that understands the symbols, associations, and meanings of social and cultural colors. Third, the analysis and interpretation of visual symbols that can express the meaning of visual symbols so that consumers can clearly understand the meaning of visual symbols and take action, and reduce the communication errors that may occur in the process of conveying meaning. According to the design strategy required. The results of this study will confirm the importance of conveying meaning through visual symbols in package design of health functional food multivitamin, and will be used as a basis for strategic research on package design that can bring about effective communication between brands and consumers. can be done.

Investigative Activities for Statistics Learning with 1st Grade Portuguese Students

The relevance of data exploration in mathematics programmes at all levels of education led us to develop a learning experience with 1st grade students with the aim of investigating the contribution of investigative activities in the learning of statistics. To achieve this objective, the following research questions were outlined: How do students implement investigative activities when learning statistics? What difficulties do students have in carrying out investigative activities when learning statistics? Adopting a qualitative and interpretative approach, data were collected through students’ written records, the reflections of the teacher, and a written test applied before and after the learning experience. The results obtained suggest that students were able to carry out investigative activities when learning statistics although with some weaknesses, namely defining the topic and research questions, and analysing the data due to the greater complexity of the research proposed by them. Students were able to formulate conclusions, although in some groups the conclusions were underdeveloped. Students were most confident in data collection, organization, and representation activities, with limitations typical of their young age (e.g., written communication). Students also showed difficulties in managing group work and knowledge of statistics, notably concerning strategies of data organization and representation and statistical measures.

An user friendly Scheme of Numerical Representation for Music Chords

The standard western music notation for the major and minor chords is avail- able in the world. But the understanding of major and minor chords is very difficult for learning peoples particularly children’s. It is very complicated to memory of all major and minor chords and also roots of all chords. This inven- tion is easily understand by children and easily captured in standard piano and other instruments. The new method of conveying chords by numerical tech- niques is used in this invention. The octave notes are derived in to two different heads of major and minor chords by numerical method. There are two main equations are implemented and it is very easy to understanding of all major and minor chords.

Features of and representational strategies in instructional videos for primary science classes

ABSTRACT Utilisation of instructional videos for science teaching has become more widespread due to the expansion of online teaching and learning environments and growing awareness of benefits of videos, such as enabling use of effective multiple representations. With this in mind, this study aimed to examine features of instructional videos for teaching scientific inquiry, a key element of science education, and learners’ engagement, a crucial issue in instruction in terms of representational strategies used. We analysed 16 instructional videos for science teaching generated by pre-service teachers. We found that the instructional videos tended to focus on posing a question related to a phenomenon and constructing its explanation conceptually rather than conducting investigations and interpreting the data. It was also found that there were alternations between providing relevant and conceptual resources and affording learners opportunities to answer questions verbally and visually to prompt their engagement. Various representational strategies, such as summarising, comparing, highlighting, sequencing, and presenting vivid phenomena, were also employed for better teaching scientific inquiry as a part of learners’ ongoing cognitive activities. Based on the findings, we argue that there is potential for using instructional videos for teaching science, considering representational strategies in terms of scientific inquiry and learners’ engagement.

Machine Learning and Ensemble Learning Based Method Using Online Employee Assessments to Identify and Analyze Job Satisfaction Factors

In this paper it was emphasized that machine learning techniques can achieve high performance in classification and work effectively and scalably with large data sets. The dataset used in this study was obtained from www.kaggle.com. A total of 67529 comments collected from people working at Google, Amazon, Netflix, Facebook, Apple and Microsoft were evaluated. The N-gram model is an important representation scheme in text mining. N-gram models are the unigram model (N = 1), bigram (N = 2), and trigram (N = 3). Three different weighting schemes as TP, TF, and TF-IDF, and three different weighting schemes for traditional machine learning-based analysis as N-gram model (bigram, unigram and trigram) was used. Five supervised learning algorithm was used to train models: Naive Bayes, Support Vector Machines (SVM), Logistic Regression (LR), K-Nearest Neighbor (KNN) and Random Forest (RF).