Scaling High-Quality Pairwise Link-Based Similarity Retrieval on Billion-Edge Graphs

SimRank is an attractive link-based similarity measure used in fertile fields of Web search and sociometry. However, the existing deterministic method by Kusumoto et al. [ 24 ] for retrieving SimRank does not always produce high-quality similarity results, as it fails to accurately obtain diagonal correction matrix D . Moreover, SimRank has a “connectivity trait” problem: increasing the number of paths between a pair of nodes would decrease its similarity score. The best-known remedy, SimRank++ [ 1 ], cannot completely fix this problem, since its score would still be zero if there are no common in-neighbors between two nodes. In this article, we study fast high-quality link-based similarity search on billion-scale graphs. (1) We first devise a “varied- D ” method to accurately compute SimRank in linear memory. We also aggregate duplicate computations, which reduces the time of [ 24 ] from quadratic to linear in the number of iterations. (2) We propose a novel “cosine-based” SimRank model to circumvent the “connectivity trait” problem. (3) To substantially speed up the partial-pairs “cosine-based” SimRank search on large graphs, we devise an efficient dimensionality reduction algorithm, PSR # , with guaranteed accuracy. (4) We give mathematical insights to the semantic difference between SimRank and its variant, and correct an argument in [ 24 ] that “if D is replaced by a scaled identity matrix (1-Ɣ)I, their top-K rankings will not be affected much”. (5) We propose a novel method that can accurately convert from Li et al. SimRank ~{S} to Jeh and Widom’s SimRank S . (6) We propose GSR # , a generalisation of our “cosine-based” SimRank model, to quantify pairwise similarities across two distinct graphs, unlike SimRank that would assess nodes across two graphs as completely dissimilar. Extensive experiments on various datasets demonstrate the superiority of our proposed approaches in terms of high search quality, computational efficiency, accuracy, and scalability on billion-edge graphs.

المعاقبة النحوية وتطبيقاتها الدلالية: معاقبة الأفعال أنموذجاً (The Grammatical Alternation and Its Semantic Applications: The Alternation of Verbs as A model)

يهدف هذا البحث إلى دراسة المعاقبة في الدرس النحوي والتعريف بمفهومها وبيان وسائلها، وخصائصها النحوية وتحليل أمثلتها الواردة في كتب النحاة دلالياً، وتأتي أهمّيّة هذه الظاهرة من حيث إنّ قِدمَها قِدَمُ اللغة، وهي ظاهرة سادت لغة بعض القبائل العربيّة التي احتجّ بها النحويون، وهذه الظاهرة توضّح الفرق الدلالي بين الكلمات، مثلاً: ريّا وروّى؛ فالأولى صفة والثانية اسم؛ لذا فهذه الظاهرة لم تكن اعتباطًا بل كانت مقصودة لبيان دلالات الكلمات فضلاً عن أهمّيّة هذه الظاهرة صوتيًّا، ويستخدم البحث المنهجين الوصفي والتحليلي وذلك من خلال جمع المعلومات اللازمة حول الدراسة من تعريفات وشروح متعلّقة بها، وتحليلها تحليلاً دلالياً للوصول إلى النتائج، ومن أهم النتائج التي توصل إليها البحث: أنّ الفعل اللازم قد يرد بمعنى فعل آخر متعد عن طريق التضمين، فيعمل عمل ذلك الفعل، فينصب المفعول به. وأنّ العرب قد استخدمت الفعل الماضي بمعنى المضارع، فيجيء بلفظ الماضي والمعنى بلفظ المضارع، وذلك أنّه أراد الاحتياط للمعنى، فجاء بمعنى المضارع المشكوك في وقوعه بلفظ الماضي المقطوع بكونه، حتى كأن هذا قد وقع واستقر لأنّه متوقع مترقب. الكلمات المفتاحية: المعاقبة النحوية، الدلالة، الأفعال. Abstract This research aims to study the alternation in the grammatical lesson, define its concept, explain its means, and grammatical characteristics, and analyze its examples in the grammarians’ books semanticly. The semantic difference between the words, for example: raya and rawa; The first is an adjective and the second is a noun. Therefore, this phenomenon was not arbitrary, but rather was intended to show the semantics of the words as well as the importance of this phenomenon phonetically. To it search: that the necessary verb may be received: The imperative verb may be received in the sense of another transitive verb by means of an inclusion, so the action of that verb works, and the object is made in it. And that the Arabs have used the past verb in the sense of the present, so he comes with the wording of the past and the meaning with the wording of the present, and that is because he wanted to be cautious of the meaning, so he came with the meaning of the present whose occurrence is doubtful with the wording of the past that is cut off from being, even as if this had occurred and settled because it is expected and awaited. Keywords: Grammatical Alternation, Semantics, Verbs.

Gender agreement with exclusive disjunction in Slovenian

This paper addresses two issues: 1. Empirically, we report novel experimental data on agreement with exclusively disjoined subjects in Slovenian; 2. Theoretically, we look into the nature of attested agreement strategies with coordinated NPs. In particular, we investigate how these strategies behave under coordinators with different semantics, i.e. exclusive disjunction and conjunction. Based on the elicitation results, we argue that closest conjunct agreement, resolved agreement, and highest conjunct agreement are all present under exclusive disjunction to different extents, which suggests a uniform set of agreement strategies under disjunction and conjunction despite the semantic difference. Further, we argue against the presence of default agreement under both disjunction and conjunction in Slovenian, and argue for a particular set of gender resolution rules.

A Study on Appearance Acceptance Appraisal of Elderly Mobility Assists

Given the low birth rate and the rapid aging of the population in Taiwan, the issues such as long-term care, high dependency ratio, and labor shortage are no longer avertible. The mobility assists shall play a significant role in helping the elder community to live independently with dignity, and the appearance of the mobility assists should make them feel comfortable as well. Hence this study recommends the development of assistive devices, barrier-free environments, and other solutions that allows the elderly to remain independent. This study explores the influencing factors of the willingness to use exoskeleton mobility assists on the psychological aspects. The purpose of this study is to explore the key factors of the willingness to use Wearable Robotic Exoskeleton. By sorting the style of the assistive devices based on various appreciation levels, the semantic difference scale was used to investigate the influence on the appearance of the assistive device on the wearing acceptance. Lastly, this study aims to acquire the factors that affect the style of the model and the influencing factors on the wearing acceptance, providing a reference for future assistive device design and developers to increase user’s willingness to accept the product.

Career Awareness in the Study of Malay Language Program

Awareness of the career to be pursued can be present as early as childhood and can vary according to knowledge and interest. Most of today's children are more focused on achieving excellence in education, but still vague in the career that will be pursued. This situation persisted when they continued their studies at higher education and they realized that the choice of field of study was not in line with their ambitions. Usually, these children choose a field based on their results rather than settling for the required degree for the field of interest. Therefore, this study was conducted to determine the extent of students’ understanding of the scope and career prospects in Malay language. Students were distributed with a set of questionnaires containing 54 items using ordinal scales of semantic difference types. The findings of the study showed that 34(56.7%) of the students studied chose this field of study as the main choice. Students also showed a high level of knowledge about careers in the Malay language which they studied (M=3.92; SD=0.57). Students have also listed a total of 14 careers in the field of Malay language they are familiar with.

MF2ResU-Net: A Multi-Feature Fusion Deep Learning Architecture for Retinal Blood Vessel Segmentation

Segmentation of blood vessels becomes an essential step in computer aided diagnosis system for the diseases in several departments of ophthalmology, neurosurgery, oncology, cardiology, and laryngology. Aiming at the problem of insufficient segmentation of small blood vessels by existing methods, a novel method based on multi-module fusion residual neural network model (MF2ResU-Net) was proposed. In the proposed networks, to obtain refined features of vessels, three cascade connected U-Net networks were employed as main networks. To deal with the problem of over-fitting, residual paths were used in main networks. In the blocks of U-Net in MF2ResU-Net, in order to remove the semantic difference in low-level and high-level, shortcut connections were used to combine the encoder layers and decoder layers in the blocks. Furthermore, atrous spatial pyramid pooling was embedded between the encoder and decoder to achieve multi-scale feature of blood vessels. During the training of the networks, to deal with the imbalance between background and foreground, a novel joint loss function was proposed based on the dice and cost- sensitive, which could greatly reduce the impact of unbalance in classes of samples. In experiment section, two retinal datasets, DRIVE and CHASE DB1, were used to test our method, and experiments showed that MF2ResU-Net was superior to existing methods on the criteria of sensitivity (Sen), specificity (Spe), accuracy (Acc), and area under curve (AUC), the values of which are 0.8013 and 0.8102, 0.9842 and 0.9809, 0.9700 and 0.9776, and 0.9797 and 0.9837 respectively for DRIVE and CHASE DB1. The results of experiments demonstrated the effectiveness and robustness of the model in the segmentation of complex curvature and small blood vessels.

Digital, analog, electronic or virtual: Which is right?

The library science terminology related to the new information technologies has become unstable with the emergence of the term “digital” as an alternative to the term “electronic”. The author expresses high opinion of the analysis of these terms, accomplished by Ksenia Yu. Volkova and Victor V. Zverevich in their article “Digital or electronic” (Nauch. and tech. b-ki. - 2020. - № 12. -P. 159-172), and as a follow-up, digs into the semantic difference of the terms content. In the professional literature, the discussion of modern terminology began with the term "virtual" in the 1990s. It was at the same time that the short life for the term was predicted, and experience speaks to the truth of the prediction. Electronic and digital document types exist in the technodocument cluster and - further - only of its technotronic subtype. The technotronic document is a document which creation and reproduction require modern technical instruments and technologies. Electronic document is the most modern type of technotronic document for the moment. The digital - electronic contraposing, which is widely used in the modern world library literature, is incorrect. In fact, the digital opposes the analog. Electronic is a general term applied both to digital and analog, as well as to virtual. The terms electronic and non-electronic make the dichotomous pair of concepts. The author foresees long relevance of the concept of electronic library science.

Coś innego to coś innego niż co innego. O tym i rzeczach podobnych

Coś innego (Something Else) Is Different than co innego (Something Different). These and Similar Issues The article discusses the opposition between Polish pronouns ending with -ś and corresponding pronouns which lack this exponent. Such an opposition can be noticed in the combinations of units of both types with the expressions: inny/indziej (else/other), (e.g. ktoś inny : kto inny, coś innego : co innego, gdzieś indziej : gdzie indziej, kiedyś indziej : kiedy indziej, etc.), as well as in some modal contexts (e.g. Potrzebuję kogoś do pomocy vs. Potrzebuję kogo do pomocy, i. e. ‘I need someone to help me’). The author of the article claims that pronouns in such pairs do not possess the same meaning, as they differ semantically and, in some cases, also pragmatically. The semantic difference between them comes down to the key contrast between the subject being spoken of and what is said about the subject. In the contexts under consideration, the pronouns ending with -ś are indexes used to denote objects in the world, while those belonging to the other type only pretend to occupy the positions of the predicate arguments, which are neither specified nor implemented. The article also considers the possibility of treating the analyzed types of expressions not as sets of separate lexemes, but rather as products of adequate operations that should be properly described.

Regression Loss in Transformer-based Supervised Neural Machine Translation

Transformer-based model has achieved human-level performance in supervised neural machine translation (SNMT), much better than the models based on recurrent neural networks (RNNs) or convolutional neural network (CNN). The original Transformer-based model is trained through maximum likelihood estimation (MLE), which regards the machine translation task as a multilabel classification problem and takes the sum of the cross entropy loss of all the target tokens as the loss function. However, this model assumes that token generation is partially independent, without realizing that tokens are the components of a sequence. To solve the problem, this paper proposes a semantic regression loss for Transformer training, treating the generated sequence as a global. Upon finding that the semantic difference is proportional to candidate-reference distance, the authors considered the machine translation problem as a multi-task problem, and took the linear combination of cross entropy loss and semantic regression loss as the overall loss function. The semantic regression loss was proved to significantly enhance SNMT performance, with a slight reduction in convergence speed.