Contributions in Merlin

We perform all possible removals of nodes from networks of size , then we identify and measure the largest connected component left in every case. The smallest of these components represents the maximum possible damage (on a network of vertices), limited to the removal of nodes, and the set that produces such damage is called the optimal set of size . We apply the procedure in a series of networks with controlled and varied modularity. Then, we compare the resulting statistics with the effect of removing the same amount of vertices according to state of the art methods of network fragmentation, i.e., High Betweenness Adaptive attack, Collective Influence, and Module-Based Attack. For practical matters we performed mainly attacks of size on networks of size , because the number of all possible sets () is at the verge of the computational capability of standard desktops. The results show, in general, that the resilience of networks to attacks has an inverse relationship with modularity, with being the critical value, from which the damage of the optimal attack increases rapidly. Networks are highly vulnerable to targeted attacks when the modularity is greater than the critical value of each heuristic method. On the other hand, for modularities lower than , all the heuristic strategies studied have a similar performance to a random attack.

The used car market is full of mistrust and uncertainties. Providing a vehicle history with trusted car data increases market transparency but also threatens the market position of used car dealers. But does a vehicle history also provide opportunities for them? Based on interviews with used car dealers, we propose three prototypes addressing current problems. The evaluation results of these prototypes indicate that they could improve the current market situation by easing data access , increasing the efficiency of sales related processes, and by expanding the competences of used car dealers. This paper is one of only few papers analyzing the current situation of used car dealers in the light of emerging new technologies. It indicates that used car dealers do not necessarily become disintermediated by the provision of trusted car data but rather have opportunities to reinvent themselves.

Carbon taxation has been studied primarily in social planner or infinitely lived agent models, which trade off the welfare of future and current generations. Such frameworks obscure the potential for carbon taxation to produce a generational win-win. This paper develops a large-scale, dynamic 55-period, OLG model to calculate the carbon tax policy delivering the highest uniform welfare gain to all generations. The OLG framework, with its selfish generations, seems far more natural for studying climate damage. Our model features coal, oil, and gas, each extracted subject to increasing costs, a clean energy sector, technical and demographic change, and Nordhaus (2017)’s temperature/damage functions. Our model’s optimal uniform welfare increasing (UWI) carbon tax starts at $30 tax, rises annually at 1.5 percent and raises the welfare of all current and future generations by 0.73 percent on a consumption-equivalent basis. Sharing efficiency gains evenly requires, however, taxing future generations by as much as 8.1 percent and subsidizing early generations by as much as 1.2 percent of lifetime consumption. Without such redistribution (the Nordhaus “optimum”), the carbon tax constitutes a win-lose policy with current generations experiencing an up to 0.84 percent welfare loss and future generations experiencing an up to 7.54 percent welfare gain. With a six-times larger damage function, the optimal UWI initial carbon tax is $70, again rising annually at 1.5 percent. This policy raises all generations’ welfare by almost 5 percent. However, doing so requires levying taxes on and giving transfers to future and current generations ranging up to 50.1 percent and 10.3 percent of their lifetime consumption. Delaying carbon policy, for 20 years, reduces efficiency gains roughly in half.

Measuring congruence between two texts has several useful applications, such as detecting the prevalent deceptive and misleading news headlines on the web. Many works have proposed machine learning based solutions such as text similarity between the headline and body text to detect the incongruence. Text similarity based methods fail to perform well due to different inherent challenges such as relative length mismatch between the news headline and its body content and non-overlapping vocabulary. On the other hand, more recent works that use headline guided attention to learn a headline derived contextual representation of the news body also result in convoluting overall representation due to the news body's lengthiness. This paper proposes a method that uses inter-mutual attention-based semantic matching between the original and synthetically generated headlines, which utilizes the difference between all pairs of word embeddings of words involved. The paper also investigates two more variations of our method, which use concatenation and dot-products of word embeddings of the words of original and synthetic headlines. We observe that the proposed method outperforms prior-arts significantly for two publicly available datasets.

Climate change communication in the mass media and other textual sources may affect and shape public perception. Extracting climate change information from these sources is an important task, e.g., for filtering content and e-discovery, sentiment analysis, automatic summarization, question-answering, and fact-checking. However, automating this process is a challenge, as climate change is a complex, fastmoving, and often ambiguous topic with scarce resources for popular text-based AI tasks. In this paper, we introduce CLIMATEXT, a dataset for sentence-based climate change topic detection, which we make publicly available. We explore different approaches to identify the climate change topic in various text sources. We find that popular keyword-based models are not adequate for such a complex and evolving task. Context-based algorithms like BERT [1] can detect, in addition to many trivial cases, a variety of complex and implicit topic patterns. Nevertheless, our analysis reveals a great potential for improvement in several directions, such as, e.g., capturing the discussion on indirect effects of climate change. Hence, we hope this work can serve as a good starting point for further research on this topic.

The digitally enabled sharing economy, also called the “digital sharing economy” (DSE), has changed patterns of consumption by introducing new choices and channels for provision and receipt of services. The DSE encompasses sharing systems whose business models may vary distinctly from platform to platform. Although business models in the context of the sharing economy have been studied so far, we have observed that the current literature does not provide an approach that covers all the possible business models (in the broadest sense of the term) that (potentially) exist within the scope of the DSE. The present paper, therefore, aims to propose a typology of business models in the DSE that covers a wide space of models – even those which may not involve “business” in the commercial sense. This is achieved through an iterative inductive process based on a design science research approach. The typology can assist in positioning the current and future sharing systems in the DSE by systematically classifying their business models. It is intended to serve as a guiding tool for the sustainability assessment of platforms from both resource and socio-economic perspectives. The present study can also enable researchers and practitioners to capture and systematically analyse digital sharing business models based on a structured, actionable approach.

We argue that incentives to take equity risk (”equity incentives”) only partially capture incentives to take asset risk (“asset incentives”). This is because leverage, while central to the theory of risk-shifting, is not explicitly considered by equity incentives. Employing measures of asset incentives that account for leverage, we find that asset risk-taking incentives can be large compared to incentives to increase firm value. Stock holdings can induce substantial risk-taking incentives, contrary to the assumption that only stock options drive risk-taking. Finally, asset incentives help explain asset risk-taking of U.S. financial institutions before the 2007/08 crisis.

This paper studies the dynamic construction of a blockchain by competitive miners. In contrast to the literature, we assume a finite time horizon. Moreover, miners are rewarded for blocks that eventually become part of the longest chain. It is shown that popular mining strategies such as adherence to conservative mining or to the longest-chain rule constitute pure-strategy Nash equilibria. However, these equilibria are not subgame perfect.

Taking advantage of football’s popularity, a field experiment was conducted on immigrants’ access to social activities. The experiment consisted of posing as football fans applying to take part in a trial training session for an amateur team and sending emails to coaches at over twenty thousand clubs all over Europe. Despite the applications being identical, those sent as if from footballers with local names obtained more affirmative responses than those sent with names of a foreign origin. In the case of Spain, the preference in favour of native players was evident: the difference was of thirteen points, standing above the European average. The results reveal difficulties for social integration that, subsequently, may also lead to barriers relating to labour and economic integration.

Das Virus fordert nicht nur die Medizin heraus: Auch die Geisteswissenschaften stehen vor neuen Fragen. Linguisten und Psychologen, Theologen und Juristen setzen sich mit den Pandemie-Folgen auseinander. Wir stellen die spannendsten Projekte vor.