Multilingual Offensive Language Identification for Low-resource Languages

Offensive content is pervasive in social media and a reason for concern to companies and government organizations. Several studies have been recently published investigating methods to detect the various forms of such content (e.g., hate speech, cyberbullying, and cyberaggression). The clear majority of these studies deal with English partially because most annotated datasets available contain English data. In this article, we take advantage of available English datasets by applying cross-lingual contextual word embeddings and transfer learning to make predictions in low-resource languages. We project predictions on comparable data in Arabic, Bengali, Danish, Greek, Hindi, Spanish, and Turkish. We report results of 0.8415 F1 macro for Bengali in TRAC-2 shared task [23], 0.8532 F1 macro for Danish and 0.8701 F1 macro for Greek in OffensEval 2020 [58], 0.8568 F1 macro for Hindi in HASOC 2019 shared task [27], and 0.7513 F1 macro for Spanish in in SemEval-2019 Task 5 (HatEval) [7], showing that our approach compares favorably to the best systems submitted to recent shared tasks on these three languages. Additionally, we report competitive performance on Arabic and Turkish using the training and development sets of OffensEval 2020 shared task. The results for all languages confirm the robustness of cross-lingual contextual embeddings and transfer learning for this task.

Determinants of Belief – And Unbelief – In Climate Change

Climate change is a most serious challenge. Committing the needed resources requires that a clear majority of citizens approves the appropriate policies, since committing resources necessarily involve a trade-off with other expenses. However, there are distinct groups of people who remain in denial about the realities of climatic change. This chapter presents a range of psychological and social phenomena that together explain the phenomena that lead to denial.

Selected Factors Determining the Adoption and Use of Participatory Budgeting in Central and Eastern Europe

This is a summary article of the SJPS thematic issue on participatory budgeting in the Central and Eastern European region. Its authors provide an overview of the diffusion of participatory budgeting, and they classify relevant countries in terms of the pace of this diffusion into four different groups: frontrunners, early majority, later majority, and lagging adopters. In addition, they uncover various diffusion mechanisms that have been used. Since the research articles included in this thematic issue unpack various factors that influence the diffusion of the innovative practice of participatory budgeting in the specific settings of Central and Eastern Europe, the main goal of this article is to sum up their crucial findings and formulate several conclusions, including a few avenues for further research. A clear majority of countries in the region have already collected a relevant amount of experience with the adoption and further use of participatory budgeting. An analysis of the individual experiences reveals that the position and characteristics of mayors, organizational resources, and available capacities, as well as the quality of public trust, are likely to be important factors that determine the adoption and use of participatory budgeting in the region.

Strict Products Liability 2.0: The Triumph of Judicial Reasoning Over Mainstream Tort Theory

Strict products liability has evolved in a manner that is widely misunderstood. The liability rule was first formulated to govern defective products that did not minimally perform one of their ordinary functions as expected by consumers—a malfunction that violates the implied warranty of quality. After adopting this rule, courts began applying it to products that did not malfunction and found that a test for defect based on consumer expectations often is indeterminate or can otherwise unduly limit liability in an important class of cases. To address these problems, most courts adopted the risk-utility test, a form of cost-benefit analysis that functions like the negligence standard of reasonable care. Relying on these cases, the Restatement (Third) of Torts: Products Liability embraced the risk-utility test, jettisoned the consumer expectations test, and characterized strict products liability as a misleading label that perpetuates confusion about liability being strict when it instead is based on negligence. In response, a clear majority of courts have rejected this negligence-based framework and affirmed the continued vitality of strict products liability. Puzzled by this unexpected development, mainstream scholars claim that courts are confused by the rhetoric of strict products liability. The prevailing scholarly opinion about this matter is confused; its fixation on negligence ignores the implied warranty rationale for strict products liability. Having been largely formulated as a rule of contract law, the implied warranty is under-theorized as a tort doctrine. Once adequately developed, the tort version of the implied warranty shows why courts have transformed the rule of strict products liability from the last century into a more comprehensive regime—“strict products liability 2.0”—that relies on consumer expectations to incorporate the risk-utility test into the framework of strict products liability. As compared to ordinary negligence liability, the implied warranty defines the safety problem in the normatively appropriate manner, thereby sharpening the inquiry about what’s at stake. In dismissing this important development, mainstream tort theory relies on legal categories that fundamentally differ from the ones courts have used to develop strict products liability with analogical reasoning. Scholars have either resorted to overly general theories of tort liability or have otherwise focused on narrow doctrinal questions. By not engaging in the mid-level categorical theorizing required by analogical reasoning, the mainstream position could not see how this characteristic form of judicial reasoning has created the substantively sound regime of strict products liability 2.0.

Lpnet: Reconstructing Phylogenetic Networks from Distances Using Integer Linear Programming

We present Lpnet, a variant of the widely used Neighbor-net method that approximates pairwise distances between taxa by a circular phylogenetic network. We use integer linear programming to replace a heurisristic part of the agglomeration procedure based on local information by an exact global solution. This approach achieves an improved approximation of the input distance for the clear majority of experiments that we have run for simulated and real data. We release an implementation in R that can handle up to 94 taxa and usually needs about one minute on a standard computer for 80 taxa.

Language, Power and Success: Bestselling Translations in the Dutch CPNB Top 100 archief

This paper investigates origins, original languages and authors of bestselling translations on the annual Dutch Top 100 bestseller list. Considering the first fifty entries on the lists from the period between 1997 and 2019, the study aims to determine the Dutch position within the World Language System. The results show that about half of all the books surveyed are translations. These come from fifteen different source languages, although a clear majority are translations from English (73.2%). The analysis confirms the notion of a World Language System with central, semi-peripheral and peripheral languages and places Dutch among the peripheral languages. Furthermore, the study reveals strong globalisation and commercialisation tendencies in the Dutch book market.

Elective Despotism

This chapter highlights India’s elective despotism. It begins by explaining ‘resort politics’. This is the Indian ritual of herding away lawmakers like rustled sheep to secure places, usually hotels and holiday resorts, to ‘protect’ them from rival parties. This usually occurs when no one party or alliance of multiple parties has a clear majority in the legislature. At such moments, lawmakers are bought for billions of rupees and promises of high office by the highest bidder. If electoral democracy is disfigured by criminality, organized violence, and chremacracy, ‘resort politics’ marks the final stage of its decadence. As parties and legislatures kowtow to political bosses and executive supremacy, the drift towards what Thomas Jefferson first called ‘elective despotism’—elected governments that concentrate power in a few cunning and bossy hands—is unmistakable. The legislatures in effect rubber-stamp executive decisions. Governments cease to be answerable to Parliament.

Part IV Applicable Standards and Forms of Emergency Measures, 8 The Form of Relief in Emergency Arbitration

This chapter studies the form of relief in emergency arbitration. In any event, a prospective applicant to an emergency arbitration will almost certainly already have in mind the form of relief they require to preserve their rights pending constitution of the arbitral tribunal. Both for the applicant, and for the respondent, the central issue when considering the form of relief in emergency arbitration will be whether the emergency arbitrator has the power to impose the requested relief. Evidently, the form of relief the emergency arbitrator is empowered to impose is classified differently across the Emergency Arbitration Rules, with no clear majority approach. The forms of relief range from "interim measures" and "interim and conservatory measures"; to "urgent measures", "urgent provisional measures", and "urgent interim and conservatory measures"; and even to "emergency measures". The different terms used to describe the types of relief available under Emergency Arbitration Rules reflects the general approach taken to urgent relief in international arbitration.

The August 1910 Elections

The August 1910 elections, for a doubled sized revisionary assembly, was the basis for change, provided the leader could resolve differences among the 'new men' and channel new ideas into practical politics. The contest was between the old political world, palaiokommatismos, built around old and powerful families (Theotokis, Rallis, etc) sometimes labelled oligarchy, and new men, independents standing for some version of revival (anorthosis). The old parties won a clear majority of seats, but the independents, new men, with about one third of the seats, had made a compelling entry to parliamentary politics and could not be ignored. They were themselves divided into groups of liberals, socialists, agrarians, republicans etc. They needed someone to bring them together and lead them, i.e. Venizelos, and a structure and organization, i.e. a political party. The aim must be to channel their energy and ideas into practical politics. Venizelos wound up his affairs in Crete and moved to the Athenian stage.

Communities Moving Ahead, Falling Behind: Evidence from the Index of Deep Disadvantage

Using a multidimensional index weighting factors related to income, health, and social mobility—the Index of Deep Disadvantage (IDD)—we rank the well-being of disadvantaged U.S. counties (initial scores below the median IDD) when they were on the cusp of the Great Recession and then again well into the recovery. We compare the characteristics of counties that saw improvements to those that saw declines. We find that a clear majority of counties were stable in relative rank. Counties showing improvement tended to have been worse off prerecession than counties where well-being declined. Improving counties were less likely to be urban, tended to have smaller fractions of the population identifying as Black and larger fractions as white, and had a lower proportion of jobs in manufacturing. Stable counties were, on average, the worst off pre-recession and thus remained the worst off near the end of the recovery. All county groups improved in income and employment through the recovery, but these advances were not consistently associated with gains in other areas such as incidence of low-weight births.