Tracking Knowledge Propagation Across Wikipedia Languages

Linking Data and Descriptions on Moths Using the Wikimedia Ecosystem

Wikipedia is one of the most widely accessed sources of information worldwide, playing a crucial role in global knowledge dissemination across languages and cultures. Understanding how information complexity varies among different language editions is essential to identify gaps and promote equitable access to knowledge. This study investigates differences in the informational complexity of Wikipedia articles across three language editions: English, Spanish and Portuguese. By analysing key quantitative features such as word count, number of topics, figures and references in articles from five thematic categories, we provide empirical evidence of significant variation in content depth. Our results reveal that both language and thematic category significantly affect Wikipedia article complexity across variables. In addition, a significant interaction between language and category was observed for number of topics. These findings highlight the linguistic disparities in knowledge representation on Wikipedia and underscore the importance of promoting linguistic diversity and inclusivity in global knowledge dissemination.

The English language Wikipedia is notable for its large number of articles. However, 288 other active language editions of Wikipedia have also developed through the intricate interactions of contributing editors. While the editor interactions in the English Wikipedia have been researched extensively, these other language editions remain understudied. To understand how editors currently come to consensus in article building in the Spanish language, a team of researchers has leveraged an existing English framework that depicts how power and policies play a role in mass collaboration. Using this English language framework, we are utilizing qualitative coding methods to build a unique model of the editor interactions on the Spanish language Wikipedia (ES). The results of this study will help contribute to a deeper understanding of how a framework in a different language edition of Wikipedia varies from the English. Our preliminary findings show that the power plays utilized in the Spanish Wikipedia talk pages differ in type and amount from power plays identified in the talk pages in the English language Wikipedia (EN), suggesting different forms of editor interactions and facilitation strategies across this multilingual platform.

Abstract. The Extended Column Test (ECT) and the Propagation Saw Test (PST) are two commonly used tests to assess the likelihood of crack propagation in a snowpack. Guidelines suggest beams with lengths of around 1 m, yet little is known about how test length affects propagation. Thus, we performed 163 ECTs and PSTs 1.0–10.0 m long. On days with full crack propagation in 1.0–1.5 m tests, we then made videos of tests 2.0–10.0 m long. We inserted markers for particle tracking to measure collapse amplitude, propagation speed, and wavelength. We also used a finite element (FE) model to simulate the strain energy release rate at fixed crack lengths. We find that (1) the proportion of tests with full propagation decreased with test length; (2) collapse was greater at the ends of the beams than in the centers; (3) collapse amplitude was independent of beam length and did not reach a constant value; (4) collapse wavelengths in the longer tests were around 3 m, two times greater than what is predicted by the anticrack model. We also confirmed the prediction that centered PSTs had double the critical length of edge PSTs. Based on our results, we conclude that cracks propagated more frequently in the shorter tests because of increased stress concentration from the far edge. The FE model suggests this edge effect occurs for PSTs of up to 2 m long or a crack to beam length ratio ≥ 0.20. Our results suggest that ECT and PST length guidelines may need to be revisited.

Linear analyses of wave-CISK models often lead to the conclusion that the mechanism does not explain the scale selection of planetary-scale tropical motions. In particular, Kelvin or gravity wave-CISK tends to prefer cumulus-scale motions. On the other hand, numerical models with “positive-only” heating did not experience much difficulty in generating large scale disturbances of realistic structure. A previous study explained these numerical results in terms of linear Kelvin wave-CISK modes, whose wave packages exhibit remarkable resemblence to the simulated disturbances. However, the linear theory can not explain how the simulated disturbances can maintain a “wavenumber-l” structure against the explosive growth of short wavelength components. A reexamination of the wave-CISK theories is carried out in this paper. It is argued that the CISK mechanism possess inherently a very severe form of nonlinearity that takes full effect for even infinitesimal perturbations. The linearized CISK theories, therefore, do not give a correct description of a CISK-driven system at any stage of its development. A simple 5-level 64-wave spectral model demonstrates that the nonlinear effects of wave-CISK alone is able to produce exponentially growing “wavenumber-1” flow patterns that propagate without change of shape. Since these flow patterns have well-defined structure, growth rate, and propagation speed, it is proposed to call them nonlinear wave-CISK modes. The nonlinear wave-CISK modes exhibit all the characteristics of the linear Kelvin wave-CISK packages of Chang and Lim (1988). The scale selection problem of their linear Kelvin wave-CISK theory has, however, been resolved. Introduction of a diffusion damping has no significant effect on the propagation speed and structure of these modes. The overall structure of the modes depends mainly on the growth rate. Varying other parameters may affect the growth rate; but modes of the same growth rate look rather like each other, irrespective of the other parameters.

With the increasing interest in location-based services, location comparison gains more and more attentions. One of the best ways to represent a location is to use topics that are generated near the location. In order to compare locations through such geographical topics, two conditions need to be met. One is that the topic set should be fixed but cover various aspects of all possible locations, and the other is that geographical topics often depend on each other. This paper proposes Probabilistic Explicit Semantic Analysis (PESA) that meets these conditions. PESA represents a location as a weighted topic vector where each topic is a Wikipedia concept. The number of Wikipedia concepts is fixed, but their enormous quantity allows PESA to be used to compare various locations. In addition, link information within Wikipedia articles is used to compute prior probabilities of topics considering their dependencies. That is, it enables PESA to model the topic dependency. PESA was evaluated using eighteen locations in three distinct geographical categories and compare it with LDA and ESA. The experimental results that PESA outperformed LDA and ESA highlighting its superiority in location comparison.

Software repositories contain a vast wealth of information about software development. Mining these repositories has proven useful for detecting patterns in software development, testing hypotheses for new software engineering approaches, etc. Specifically, mining source code has yielded significant insights into software development artifacts and processes. Unfortunately, mining source code at a large-scale remains a difficult task. Previous approaches had to either limit the scope of the projects studied, limit the scope of the mining task to be more coarse-grained, or sacrifice studying the history of the code due to both human and computational scalability issues. In this paper we address the substantial challenges of mining source code: a) at a very large scale; b) at a fine-grained level of detail; and c) with full history information.

The Hawai‘i Natural Energy Institute (HNEI) is leading a team engaged in the research, development, deployment, and analysis of grid-scale battery energy storage systems (BESS) that are designed for system control and power quality support at the generation, transmission, and distribution levels. The program aims to identify high value BESS applications at various system levels, develop control algorithms that maximize the benefit to the grid/customer and the lifetime of the BESS, and evaluate and optimize those algorithms under real world operating conditions. The focus is to deploy, operate, and validate the performance of four grid-scale BESS for various ancillary service applications on grid systems across the state. Large scale battery energy storage systems will become an important part of the electric grid in the near future and it will be essential to ensure their reliability. The objective of this project is to understand the degradation of the individual batteries to anticipate failures. Laboratory testing of advanced Li-ion battery cells is performed to support life-time analysis of technologies targeted for large-scale grid energy storage applications. HNEI’s battery testing efforts have focused on grid scale deployment lithium ion titanate battery technology which is successfully being used for a variety of purposes including frequency and power regulation of large scale wind and solar energy generation. There is currently a lack of understanding of long term performance of battery technology in general, and specifically lithium ion titanate, used under large scale, grid conditions. This work aims to fill that gap in knowledge through laboratory scale battery testing and the development predictive lifetime models of performance of grid battery technology. Accelerated testing of identical lithium ion titanate battery technology was performed in the laboratory and those results will be used to develop predictive performance models. This model will combine existing modules [1-4] that account for single cell asymmetric degradation, single cell heat generation, string imbalance, and cell paralleling. As real world data is collected from the grid batteries, the predictive models will be compared and assessed for accuracy and ability to predict performance. This work will present preliminary results of the cycle and calendar aging long-term studies. [1] M. Dubarry, N. Vuillaume and B. Y. Liaw, J. Power Sources 186(2), (2009) 500-507. [2] M. Dubarry, C. Truchot and B.Y. Liaw, J. Power Sources 219 (2012) 204-216. [3] https://www.soest.hawaii.edu/HNEI/alawa/[4] M. Dubarry, C. Truchot, A. Devie and B. Y. Liaw, J. Electrochem. Soc 162(6), (2015) A877-A884.

We demonstrate how crowdsourcing can be used to automatically build a personalized tourist attraction recommender system, which tailors recommendations to specific individuals, so different people who use the system each get their own list of recommendations, appropriate to their own traits. Recommender systems crucially depend on the availability of reliable and large scale data that allows predicting how a new individual is likely to rate items from the catalog of possible items to recommend. We show how to automate the process of generating this data using crowdsourcing, so that such a system can be built even when such a dataset is not initially available. We first find possible tourist attractions to recommend by scraping such information from Wikipedia. Next, we use crowdsourced workers to filter the data, then provide their opinions regarding these items. Finally, we use machine learning methods to predict how new individuals are likely to rate each attraction, and recommend the items with the highest predicted ratings.

A system of hydrodynamic equations in the presence of large-scale inhomogeneities for a high plasma beta solar wind is derived. The theory is derived under the assumption of low turbulent Mach number and is developed for the flows where the usual incompressible description is not satisfactory and a full compressible treatment is too complex for any analytical studies. When the effects of compressibility are incorporated only weakly, a new description, referred to as "nearly incompressible hydrodynamics," is obtained. The nearly incompressible theory, was originally applied to homogeneous flows. However, large-scale gradients in density, pressure, temperature, etc., are typical in the solar wind and it was unclear how inhomogeneities would affect the usual incompressible and nearly incompressible descriptions. In the homogeneous case, the lowest order expansion of the fully compressible equations leads to the usual incompressible equations, followed at higher orders by the nearly incompressible equations, as introduced by Zank and Matthaeus. With this work we show that the inclusion of large-scale inhomogeneities (in this case time-independent and radially symmetric background solar wind) modifies the leading-order incompressible description of solar wind flow. We find, for example, that the divergence of velocity fluctuations is nonsolenoidal and that density fluctuations can be described to leading order as a passive scalar. Locally (for small lengthscales), this system of equations converges to the usual incompressible equations and we therefore use the term "locally incompressible" to describe the equations. This term should be distinguished from the term "nearly incompressible," which is reserved for higher-order corrections. Furthermore, we find that density fluctuations scale with Mach number linearly, in contrast to the original homogeneous nearly incompressible theory, in which density fluctuations scale with the square of Mach number. Inhomogeneous nearly incompressible equations for higher order fluctuation components are derived and it is shown that they converge to the usual homogeneous nearly incompressible equations in the limit of no large-scale background. We use a time and length scale separation procedure to obtain wave equations for the acoustic pressure and velocity perturbations propagating on fast-time-short-wavelength scales. On these scales, the pseudosound relation, used to relate density and pressure fluctuations, is also obtained. In both cases, the speed of propagation (sound speed) depends on background variables and therefore varies spatially. For slow-time scales, a simple pseudosound relation cannot be obtained and density and pressure fluctuations are implicitly related through a relation which can be solved only numerically. Subject to some simplifications, a generalized inhomogeneous pseudosound relation is derived. With this paper, we extend the theory of nearly incompressible hydrodynamics to flows, including the solar wind, which include large-scale inhomogeneities (in this case radially symmetric and in equilibrium).

Wikipedia is the largest online encyclopedia, used by algorithms and web users as a central hub of reliable information on the web. The quality and reliability of Wikipedia content is maintained by a community of volunteer editors. Machine learning and information retrieval algorithms could help scale up editors' manual efforts around Wikipedia content reliability. However, there is a lack of large-scale data to support the development of such research. To fill this gap, in this paper, we propose Wiki-Reliability, the first dataset of English Wikipedia articles annotated with a wide set of content reliability issues. To build this dataset, we rely on Wikipedia "templates". Templates are tags used by expert Wikipedia editors to indicate content issues, such as the presence of "non-neutral point of view" or "contradictory articles", and serve as a strong signal for detecting reliability issues in a revision. We select the 10 most popular reliability-related templates on Wikipedia, and propose an effective method to label almost 1M samples of Wikipedia article revisions as positive or negative with respect to each template. Each positive/negative example in the dataset comes with the full article text and 20 features from the revision's metadata. We provide an overview of the possible downstream tasks enabled by such data, and show that Wiki-Reliability can be used to train large-scale models for content reliability prediction. We release all data and code for public use.

Decentralized search in networks is an activity that is often performed in online tasks. It refers to situations where a user has no global knowledge of a network's topology, but only local knowledge. On Wikipedia for instance, humans typically have local knowledge of the links emanating from a given Wikipedia article, but no global knowledge of the entire Wikipedia graph. This makes the task of navigation to a target Wikipedia article from a given starting article an interesting problem for both humans and algorithms. As we know from previous studies, people can have very efficient decentralized search procedures that find shortest paths in many cases, using intuitions about a given network. These intuitions can be modeled as hierarchical background knowledge that people access to approximate a networks' topology. In this paper, we explore the differences and similarities between decentralized search that utilizes hierarchical background knowledge and actual human navigation in information networks. For that purpose we perform a large scale study on the Wikipedia information network with over 500,000 users and 1,500,000 click trails. As our results reveal, a decentralized search procedure based on hierarchies created directly from the link structure of the information network simulates human navigational behavior better than simulations based on hierarchies that are created from external knowledge.

Manually annotating large scale content such as Internet videos is an expensive and consuming process. Furthermore, community-provided tags lack consistency and present numerous irregularities. This chapter aims to provide a forum for the state-of-the-art research in this emerging field, with particular focus on mechanisms capable of exploiting the full range of information available online to predict user tags automatically. The exploited information covers both semantic metadata including complementary information in external resources and embedded low-level features within the multimedia content. Furthermore, this chapter presents a framework for predicting general tags from the associated textual metadata and visual features. The goal of this framework is to simplify and improve the process of tagging online videos, which are unbounded to any particular domain. In this framework, the first step is to extract named entities exploiting complementary textual resources such as Wikipedia and WordNet. To facilitate the extraction of semantically meaningful tags from a largely unstructured textual corpus, this framework employs GATE natural language processing tools. Extending the functionalities of the built-in GATE named entities, the framework also integrates a bag-of-articles algorithm for effectively extracting relevant articles from the Wikipedia articles. Experiments were conducted for validation of the framework against MediaEval 2010 Wild Wild Web dataset for the tagging task.

Understanding of fracture mechanics of the human knee structures within total knee replacement (TKR) allows a better decision support for bone fracture prevention. Numerous studies addressed these complex injuries involving the femur bones but the full macro-crack propagation from crack initiation to final failure and age-related effects on the tibia bone were not extensively studied. The present study aimed to develop a patient-specific model of the human tibia bone and the associated TKR implant, to study fatigue and fracture behaviors under physiological and pathological (i.e., age-related effect) conditions. Computed tomography (CT) data were used to develop a patient-specific computational model of the human tibia bone (cortical and cancellous) and associated implants. First, segmentation and 3D-reconstruction of the geometrical models of the tibia and implant were performed. Then, meshes were generated. The locations of crack initiation were identified using the clinical observation and the fatigue crack initiation model. Then, the propagation of the crack in the bone until final failure was investigated using the eXtended finite element method (X-FEM). Finally, the obtained outcomes were analyzed and evaluated to investigate the age-effects on the crack propagation behaviors of the bone. For fatigue crack initiation analysis, the stress amplitude–life S–N curve witnessed a decrease with increasing age. The maximal stress concentration caused by cyclic loading resulted in the weakening of the tibia bone under TKR. For fatigue crack propagation analysis, regarding simulation with the implant, the stress intensity factorand the energy release rate tended to decrease, as compared to the tibia model without the implant, from 0.152.5 to 0.111.9 (MPa) and from 10240 to 5133 (J), respectively. This led to the drop in crack propagation speed. This study provided, for the first time, a detailed view on the full crack path from crack initiation to final failure of the tibia bone within the TKR implant. The obtained outcomes also suggested that age (i.e., bone strength) also plays an important role in tibia crack and bone fracture. In perspective, patient-specific bone properties and dynamic loadings (e.g., during walking or running) are incorporated to provide objective and quantitative indicators for crack and fracture prevention, during daily activities.

Simultaneous velocity–density measurements of downslope density clouds