Large Web Research Articles

Efficient Multivariate Time Series Anomaly Detection Through Transfer Learning for Large-Scale Software Systems

Timely anomaly detection of multivariate time series (MTS) is of vital importance for managing large-scale software systems. However, many deep learning-based MTS anomaly detection models require long-term MTS training data to achieve optimal performance, which often conflicts with the frequent pattern changes observed in software systems. Moreover, the training overhead of vast MTS in large-scale software systems is unacceptably high. To address these issues, we design OmniTransfer , a model-agnostic framework that combines weighted hierarchical agglomerative clustering with an adaptive transfer learning strategy, making many state-of-the-art (SOTA) MTS anomaly detection models efficient and effective. Extensive experiments using real-world data from a large web content service provider and a network operator show that OmniTransfer significantly reduces the model initialization time by 46.49% and the training cost by 74.51%, while maintaining high accuracy in detecting anomalies.

Buckling behavior of thin-walled cold-formed steel latticed column with lacings under axial compression

Extensive research has been undertaken on built-up cold-formed steel (CFS) column with I-shaped or box-shaped section. However, there has been little attention given to lattice CFS columns, particularly regarding their buckling performance and the design methods for determining buckling capacity. This paper proposes a latticed CFS column, formed by combining two G-shaped (complex edge stiffener) channels using lacings (LCCFS). Eighteen LCCFS columns were tested under axial compression, divided into two groups with web widths of 75 mm and 200 mm. The experimental results demonstrated that no distortional buckling failure occurred in the G-shaped channels of all specimens, mainly due to the constraint provided by lacings on the flanges. A general consistency in the ultimate bearing capacities and failure modes was observed among duplicate specimens. Specimens with a small slenderness ratio predominantly experienced local buckling failure, leading to a more abrupt failure. In contrast, specimens characterized by a large slenderness ratio primarily demonstrated failure due to global buckling. The specimens with larger web width-to-thickness ratios exhibited earlier and more pronounced local buckling, resulting in a reduction in axial compressive stiffness. Furthermore, FE models have been developed to analyze the effects of different parameters on the axial compressive buckling performance of LCCFS columns. The existing direct strength method exhibited notable deviations in predicting axial compressive buckling capacity of LCCFS column, which are deemed unsatisfactory. Based on direct strength method, a method was proposed for determining the axial compressive buckling capacity of LCCFS column.

Neutrophil extracellular traps protect the kidney from ascending infection and are required for a positive leukocyte dipstick test.

Lower urinary tract infection (UTI) is common but only rarely complicated by pyelonephritis. However, the mechanisms preventing extension to the kidney are unclear. Here, we identified neutrophil extracellular traps (NETs) in healthy human urine that provide an antibacterial defense strategy within the urinary tract. In both in vivo murine models of UTI where uropathogenic E. coli are inoculated into the bladder and ex vivo human urine models, NETs interacted with uromodulin to form large webs that entrapped the bacteria. Peptidyl arginine deiminase 4 (PADI4) inhibition in mice blocked NETosis and resulted in progression of cystitis into pyelonephritis, suggesting that NETosis of urinary neutrophils acts to prevent bacterial ascent into the kidney. Analysis of UK Biobank data revealed that genetic variants in PADI4 that associated with increased risk of rheumatoid arthritis in multiple genome-wide association studies were consistently associated with reduced susceptibility to UTI. Last, we showed that urine dipstick testing for leukocyte esterase was negative in the presence of intact blood neutrophils but became positive when neutrophils were stimulated to NET, and this could be prevented by selective PADI4 inhibition, demonstrating that this test does not detect absolute neutrophil count, as has long been assumed, but specifically detects neutrophils that have undergone NETosis. These findings highlight the role of NETosis in preventing ascending infections in the urinary tract and improve our understanding of one of the most common clinical tests in medicine.

VTPL: Visual and text prompt learning for visual-language models

Visual-language (V-L) models have achieved remarkable success in learning combined visual–textual representations from large web datasets. Prompt learning, as a solution for downstream tasks, can address the forgetting of knowledge associated with fine-tuning. However, current methods focus on a single modality and fail to fully use multimodal information. This paper aims to address these limitations by proposing a novel approach called visual and text prompt learning (VTPL) to train the model and enhance both visual and text prompts. Visual prompts align visual features with text features, whereas text prompts enrich the semantic information of the text. Additionally, this paper introduces a poly-1 information noise contrastive estimation (InfoNCE) loss and a center loss to increase the interclass distance and decrease the intraclass distance. Experiments on 11 image datasets show that VTPL outperforms state-of-the-art methods, achieving 1.61%, 1.63%, 1.99%, 2.42%, and 2.87% performance boosts over CoOp for 1, 2, 4, 8, and 16 shots, respectively.

Coding Beauty and Decoding Ugliness: The Role of Aesthetic Concerns in Programming Practices

In this article, we analyze the productive role of aesthetics in organizing technoscientific work. Specifically, we investigate how aesthetic judgments form and inform code-writing practices at a large web services company in Russia. We focus on how programmers express aesthetic judgments about code and software design in everyday practice and explore how language with positive and negative valences is deployed. We find that programmers label code as “beautiful” without defining or establishing agreement about the term and are thereby able to maintain different ideals of beauty within the same organization. However, by learning how to avoid what senior developers deem to be “ugly” code, developers become socialized into producing code with a similar style and logic that we describe as “not ugly.” The fieldwork suggests that aesthetic language can function simultaneously as a mechanism that supports professional diversity within an organization and as a tool for producing consistencies in software design. Studying manifestations of both positive and negative aesthetic language in technoscientific work provides insight into professional practices and the various roles aesthetic language can play in organizational life.

Numerical Study on the Impact Response of Steel Beams with Large Web Openings: Investigating Key Parameters

Numerical Study on the Impact Response of Steel Beams with Large Web Openings: Investigating Key Parameters

GFRP continuous RC beams having web openings and externally strengthened with CFRP composites

With the enormous development of modern buildings, it is highly required to accommodate web openings in continuous reinforced concrete (RC) beams for passing utility service equipment. A wide range of studies had been conducted with regard to this subject. wtCurrent study examined numerically the major effect of the large rectangular web openings in (GFRP) reinforced concrete continuous beams under concentrated loads. Moreover, it investigated the effects of different CFRP strengthening schemes that would retrieve the ultimate capacity of the solid beam. Based on that, the main parameters that had been utilized here are the opening size, opening location and the strengthening scheme. Results showed that openings reduced the ultimate load capacity of the beams and increased the overall deflections. Furthermore, using CFRP sheets to strengthen continuous RC beams including opening retrieved (70%) of the ultimate capacity of continuous RC solid beams especially beams with openings near the external support or midspan. Likewise, using CFRP composites retrieved (48%) of the ultimate capacity of continuous RC solid beams with openings near the interior support. Furthermore, the optimum location for constructing web openings was near the external supports, which was subjected to the lowest values of shear forces and sagging moments and leaded to accommodate greater ultimate loads (85.18 kN) than other beams of other openings location.

Failure investigation of steel and brick aggregate concrete composite construction with angle and channel shear connectors

ABSTRACT This investigation focuses mainly on the failure behaviour and performance of angle and channel shear connectors in composite structures made with steel and brick aggregate concrete. It was observed that both angle and channel shear connectors did not fail themselves. However, in some specimens, localised connector yielding was observed in channel shear connectors. In both cases, concrete strata failed as either concrete crushing and shearing, combined splitting and bending, or complete splitting mode of failure. Crushing and shearing failure was noticed in specimens having smaller web and flange and were brittle in nature as compared to the failure of specimens with larger web and flange. Other parameters, such as shear resistance, energy absorption, and load-slip behaviour, were also investigated. Web height and flange width of both connectors significantly affected the ultimate shear capacity and slip of the brick aggregate concrete specimen. Channel connectors were found more flexible and carried more shear than angle connectors. Channel connectors had an average of 32.41% and 116.17% more ultimate load and slip capacity than the angle connectors in brick aggregate concrete, respectively. The load-carrying capacity, slip, and failure behaviour of the brick aggregate concrete were found similar to those of stone aggregate concrete.

Conceptualizing expectations - Vietnam’s responses during COVID-19 as an illustration

Background A key part of the complexity of actual social life lies in the fact that it operates on a large web of not only prescribed norms and rules, but also explicit and especially implicit expectations which constantly interact with each other. Methods This paper offers the first in-depth study on ‘expectations’ in bioethical and health law studies, taking Vietnam’s responses to coronavirus disease (COVID-19) as an illustration. Results It defines ‘expectations’ as normative imaginaries that can induce or guide actions and inactions at individual and collective levels. Conclusions It suggests that ‘expectations’ is a fruitful concept for bioethical and health law studies because it helps better understand the complexity of interactions in society and patterns of thinking and acting of actors in real-world contexts. Studying expectations has implications for governance and future research agendas.

Large scale dynamic web deployment for CERN, experience and outlook

CERN hosts more than 1200 websites essential for the mission of the Organization, both for internal and external collaboration and communication, as well as public outreach. The complexity and scale of CERN’s online presence is very diverse with some websites, like home.cern, accommodating more than one million unique visitors in a day. However, regardless of their diversity, all websites are created using the Drupal content management system (CMS), and are self-hosted directly in the CERN Data Center on a dedicated infrastructure that runs on Kubernetes. Workflows like provisioning, deleting, cloning, upgrading, and similar are fully automated and managed by a customized Kubernetes controller. By leveraging the custom controller, the infrastructure has proven highly reliant with minimal, manual intervention necessary. In order to further automate deployments and improve governance, a customized version of Drupal called the CERN Drupal Distribution is implemented. Supported by end-to-end integration tests and automated browser simulation, this setup enables the propagation of security and feature updates seamlessly to all websites without any downtime. This paper outlines the architecture which allows building, testing, and distributing updates to a large number of websites without any downtime. Furthermore, it presents experiences and learnings from managing such a service at CERN with limited resources.

Review of mathematical models and information technologies for business analysis of the big web data

The article provides a comprehensive review of mathematical models and information technologies used for analyzing large amounts of data in web applications. The latest re-search and publications in the field are analyzed, including a comparative analysis of ma-chine learning methods, text, image, video analysis, social network analysis, and graph algo-rithms. The goal of this research is to analyze the effectiveness and applicability of mathe-matical models and information technologies in business analysis of large web data. The arti-cle presents the results of the research and a comparative analysis of the efficiency of meth-ods, which will help business analysts choose the optimal tools for processing and analyzing large amounts of data in web applications. The article begins with an overview of the problem and the latest research and publica-tions in the field. The article provides a detailed description of various mathematical models and information technologies, including their strengths and weaknesses. A comparative analysis of these methods is presented, with a focus on their effectiveness and applicability in business analysis. The article also provides a detailed description of the applications of mathematical models and information technologies in various industries, such as e-commerce and supply chain management. The article analyzes the challenges and opportunities associated with the use of these technologies in business analysis and provides recommendations for businesses that want to take advantage of these technologies. Overall, the article provides a comprehensive overview of mathematical models and in-formation technologies used in business analysis of large web data. The article is a valuable resource for business analysts, data scientists, and researchers who want to learn more about the latest developments in this field.

Pain before, during, and after nonsuicidal self-injury: Findings from a large web study.

Competing models suggest that physical pain may play an important role in nonsuicidal self-injury (NSSI) via pain onset or pain offset, or that pain may be absent (analgesia). Few studies have tested these models in the same sample or examined factors that could explain differences in NSSI pain experience. We assessed 1,630 individuals with NSSI histories in an online survey. We descriptively examined pain during NSSI and tested preregistered hypotheses that NSSI frequency, NSSI severity, borderline personality disorder (BPD) features, emotional pain, and dissociation during NSSI are associated with experiencing less NSSI pain. Exploratorily, we also tested whether self-punishment motives were associated with less NSSI pain. Almost all participants reported recent and frequent NSSI. Participants were heterogenous in their report of NSSI pain. We found minimal support for analgesia (reported by only 4.3% of participants). More participants reported pain onset than offset, but offset was associated with reductions in emotional pain. Emotional pain was elevated prior to NSSI and decreased significantly during and after NSSI. We found that higher dissociation during NSSI was associated with less NSSI pain. Contrary to hypotheses, NSSI severity, emotional pain prior to NSSI, and self-punishment motives were associated with greater NSSI pain. NSSI frequency and BPD features were not associated with NSSI pain. BPD features interacted with dissociation and emotional pain prior to NSSI. Findings contrast with laboratory pain induction work, suggesting that, though people who self-harm may have heightened pain tolerance, they may seek to self-injure in a manner that results in pain. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Statistics of remote regions of networks

We delve into the statistical properties of regions within complex networks that are distant from vertices with high centralities, such as hubs or highly connected clusters. These remote regions play a pivotal role in shaping the asymptotic behaviours of various spreading processes and the features of associated spectra. We investigate the probability distribution P≥m(s) of the number s of vertices located at distance m or beyond from a randomly chosen vertex in an undirected network. Earlier, this distribution and its large m asymptotics 1/s2 were obtained theoretically for undirected uncorrelated networks (Dorogovtsev et al., 2003). Employing numerical simulations and analysing empirical data, we explore a wide range of real undirected networks and their models, including trees and loopy networks, and reveal that the inverse square law is valid even for networks with strong correlations. We observe this law in the networks demonstrating the small-world effect and containing vertices with degree 1 (so-called leaves or dead ends). We find the specific classes of networks for which this law is not valid. Such networks include the finite-dimensional networks and the networks embedded in finite-dimensional spaces. We notice that long chains of nodes in networks reduce the range of m for which the inverse square law can be spotted. Interestingly, we detect such long chains in the remote regions of the undirected projection of a large Web domain.

Ultimate Load Behaviour of Castellated Beams with Stiffened Octagonal Openings

ctagonal castellated beam is fabricated by adding spacer plates between the web joints of the hexagonal castellated beam to further increase the beam depth. The resulted deep beam is advantageous to enhance the shear carrying capacity and moment resistance. However, large web openings are prone to Vierendeel mechanism in the deformation that lowers the overall beam performance due to formation of plastic hinges. The present paper is concerned with non-linear analysis to predict the ultimate load behaviour of octagonal castellated beams. Finite element models were developed by using a commercial programme LUSAS. The numerical models were first validated against the experimental results reported by other researchers. This study aims to propose the ideal configuration of web opening stiffeners and evaluate the effectiveness of the additional stiffening material for different beam span lengths. Provision of ring stiffeners is effective in short span beams in respect of resistance to Vierendeel failure. The stiffeners have increased the ultimate load to the extent of 73% in which the strength restoration is considerably near to the capacity of its parent universal beam. In long span beams, however, the stiffeners can only improve the performance up to 26.61% with larger additional steel materialthan that of shorter span. In all cases, the stiffener thickness is found to be the governing factor to restoring the loss of shear strength due to large web openings. The number of stiffeners contributes no appreciable variation of ultimate load particularly in long span beams where bending is prominent

Automatic Genre Identification for Robust Enrichment of Massive Text Collections: Investigation of Classification Methods in the Era of Large Language Models

Massive text collections are the backbone of large language models, the main ingredient of the current significant progress in artificial intelligence. However, as these collections are mostly collected using automatic methods, researchers have few insights into what types of texts they consist of. Automatic genre identification is a text classification task that enriches texts with genre labels, such as promotional and legal, providing meaningful insights into the composition of these large text collections. In this paper, we evaluate machine learning approaches for the genre identification task based on their generalizability across different datasets to assess which model is the most suitable for the downstream task of enriching large web corpora with genre information. We train and test multiple fine-tuned BERT-like Transformer-based models and show that merging different genre-annotated datasets yields superior results. Moreover, we explore the zero-shot capabilities of large GPT Transformer models in this task and discuss the advantages and disadvantages of the zero-shot approach. We also publish the best-performing fine-tuned model that enables automatic genre annotation in multiple languages. In addition, to promote further research in this area, we plan to share, upon request, a new benchmark for automatic genre annotation, ensuring the non-exposure of the latest large language models.

Investigation on shear buckling of steel welded I‐section beams with reinforced web openings

AbstractPlate girders with I‐section frequently require large web openings to accommodate ducts and pipes in buildings and bridges. It is known that the critical load equivalent to an unperforated plate can be achieved by an appropriate reinforcement of the web opening. The majority of the investigations on thin‐walled webs with web opening under shear loading has been focused on web slendernesses ranging between 200 to 360. Three tests on plate girders with a web slenderness of 416 are presented that contain web openings. The effect of an all‐around reinforcement (hollow section) of the web opening is studied in three‐point bending tests. The tests are calculated with finite elements accounting for imperfections that had been measured before testing. A parametric study is performed to evaluate different types of reinforcements for high web slendernesses.

Bearing capacity of headed studs under combined tension and shear loading in composite beams with large web openings

AbstractThe use of steel‐concrete composite structures with a high installation grade is becoming increasingly popular in building and industrial construction. To reduce construction height building services are routed through web openings in the girders. However, these openings represent local points of interference where, in the case of composite girders, parts of the acting shear forces are transferred from the web of the steel profile into the concrete slab. This results in additional, local stresses in the remaining steel girder and in the concrete chord, which must be considered in the design of the floor girders and shear connectors. Web openings in composite girders are not addressed in the current European design standard for steel‐concrete composite construction (EC 4). Only with the introduction of the second generation of Eurocodes in the coming years this case will be normatively regulated. During the preparation of the standard further questions about the load‐bearing behaviour arose, as the existing design approaches usually only consider a selection of possible failure scenarios or neglect essential verification steps. A joint research project between the University of Kaiserslautern‐Landau and RWTH Aachen University aims to develop a suitable design concept for composite girders with web openings. Model and beam tests accompanied by numerical simulations will provide further information on the complex local and global load‐bearing behaviour of the composite girders.This paper presents the results of experimental investigations and numerical simulations on the influence of large web openings on the global load‐bearing behaviour of the shear connectors in the composite joint. Based on 60 pull‐out tests and numerical simulations, the load‐bearing behavior of head bolt anchors under combined shear and tensile loading was researched and evaluated.

The biological invasion of an apex predator (Silurus glanis) amplifies PCB transfer in a large lake food web

Invasive species can affect food web structure possibly modifying the transfer of pollutants in ecosystems but this facet of biological invasion remains largely unexplored. We examined how trophic and ontogenetic characteristics of the invasive European catfish could differ from its native counterpart, the Northern pike, possibly resulting in the amplification of PCB transfer to the higher trophic levels in a large lake food web. The PCB contents of catfish and pike were on average low (Ʃ7 PCBi 42.4 ± 38.6 ng g−1 ww and 37.9 ± 49.4 ng g−1 ww respectively) and dominated by PCB153 (~35 % of the PCB contamination). Only the largest pike (126 cm) slightly exceeded the European sanitary threshold of 125 ng g−1 ww Ʃ6 PCBi-NDL. Both species increased in trophic position with body size while catfish had clearly higher littoral reliance than pike indicating they exploited complementary trophic niches. PCB biomagnification was identified only for catfish (PCB153, Ʃ7 PCBi) leading to trophic magnification factor of ~5. PCB ontogenetic bioaccumulation was pervasive for catfish (PCB101, PCB118, PCB153, PCB138 and Ʃ7 PCBi) and identified for pike only regarding PCB101. The derived size accumulation factors (~1.02) indicated a size-doubling PCB contamination of ~40 cm for catfish. This finding suggested that catfish would exceed the European sanitary threshold at body size larger than 168 cm possibly constraining their commercial exploitation. Our results highlighted that the invasive catfish was a littoral-oriented apex predator occupying an alternative trophic niche as compared to pike thereby modifying the lake food web structure that resulted in an enhancement of PCB transfer to higher trophic levels. The biomagnification and ontogenetic bioaccumulation of catfish underlined the impact of this biological invasion on the fate of PCB in the ecosystem. Finally, the remarkable inter-individual PCB contamination suggested variable inter-individual PCB exposure likely associated to localized hotspots of PCB contamination in the lake.

Mining user's navigation structure by filtering impurity nodes for generating relevant predictions

Web Navigation Prediction (WNP) has been popularly used for finding future probable web pages. Obtaining relevant information from a large web is challenging, as its size is growing with every second. Web data may contain irrelevant noise, and thus cleaning such data is essential. In this paper, we emphasize the identification and elimination of noisy information from user-navigated web pages. The pruning of user browsed sessions by removing noisy webpages and their relations can help in the development of high-performing prediction models with fewer prediction errors. To minimize the prediction errors, we have proposed four pruning models namely PM3ER, PM3EN, PM3GI, and PM3EP which remove noisy web pages and their relations from the model consisting of varied simple and complex navigations. The study reveals that PM3ER is effective in prediction for websites with complicated structures resulting in complex navigations, and PM3EN is effective in predicting websites with a simple tree-like structure resulting in simple navigations. PM3ER has shown an improvement in predictions by up to 3% whereas PM3EN has attained improvement up to 5.45%.

Language contact between Italian and English: a case study on nouns ending in the suffix -ing

Abstract The article deals with how English deverbal nouns with the suffix -ing have been imported into Italian. The focus is on the semantic characteristics of these borrowed nouns in Italian and, in particular, on the question of whether they have been borrowed not as simple sign-concept pairings but with argument and event structure. In previous research, it has been claimed that argument and event structure need to be licensed by some overt functional element. Hence, borrowed deverbal nouns should have argument structure and event structure only if they have an overt affix, in other words, only if the forms are not borrowed holistically as unsegmented words but retain internal morphological structure, implying that morphological borrowing of the affix has occurred. When a foreign affix combines with native bases of the recipient language, this is often considered an important criterion for morphological borrowing, which is clearly not the case for the suffix -ing in Italian. Here, it will be shown that contrary to expectation, numerous occurrences with argument and event structure may be found in a large Italian web corpus for a sample of English deverbal nouns ending in -ing, borrowed into (certain registers) of Italian.