Model-driven engineering for mobile robotic systems: a systematic mapping study

Software and Systems Modeling - Mobile robots operate in various environments (e.g. aquatic, aerial, or terrestrial), they come in many diverse shapes and they are increasingly becoming parts of...     

Structural Investigation and Molecular Modeling Studies of Strobilurin-Based Fungicides Active against the Rice Blast Pathogen Pyricularia oryzae

The increasing emergence of fungicide-resistant pathogens requires urgent solutions for crop disease management. Here, we describe a structural investigation of new fungicides obtained by combining strobilurin and succinate dehydrogenase inhibitor pharmacophores. We identified compounds endowed with very good activity against wild-type Pyricularia oryzae, combined in some cases with promising activity against strobilurin-resistant strains. The first three-dimensional model of P. oryzae cytochrome bc1 complex containing azoxystrobin as a ligand was developed. The model was validated with a set of commercially available strobilurins, and it well explains both the resistance mechanism to strobilurins mediated by the mutation G143A and the activity of metyltetraprole against strobilurin-resistant strains. The obtained results shed light on the key recognition determinants of strobilurin-like derivatives in the cytochrome bc1 active site and will guide the further rational design of new fungicides able to overcome resistance caused by G143A mutation in the rice blast pathogen.     

A multimodal feature learning approach for sentiment analysis of social network multimedia

In this paper we investigate the use of a multimodal feature learning approach, using neural network based models such as Skip-gram and Denoising Autoencoders, to address sentiment analysis of micro-blogging content, such as Twitter short messages, that are composed by a short text and, possibly, an image. The approach used in this work is motivated by the recent advances in: i) training language models based on neural networks that have proved to be extremely efficient when dealing with web-scale text corpora, and have shown very good performances when dealing with syntactic and semantic word similarities; ii) unsupervised learning, with neural networks, of robust visual features, that are recoverable from partial observations that may be due to occlusions or noisy and heavily modified images. We propose a novel architecture that incorporates these neural networks, testing it on several standard Twitter datasets, and showing that the approach is efficient and obtains good classification results.     

Quality metrics in high-dimensional data visualization: an overview and systematization

In this paper, we present a systematization of techniques that use quality metrics to help in the visual exploration of meaningful patterns in high-dimensional data. In a number of recent papers, different quality metrics are proposed to automate the demanding search through large spaces of alternative visualizations (e.g., alternative projections or ordering), allowing the user to concentrate on the most promising visualizations suggested by the quality metrics. Over the last decade, this approach has witnessed a remarkable development but few reflections exist on how these methods are related to each other and how the approach can be developed further. For this purpose, we provide an overview of approaches that use quality metrics in high-dimensional data visualization and propose a systematization based on a thorough literature review. We carefully analyze the papers and derive a set of factors for discriminating the quality metrics, visualization techniques, and the process itself. The process is described through a reworked version of the well-known information visualization pipeline. We demonstrate the usefulness of our model by applying it to several existing approaches that use quality metrics, and we provide reflections on implications of our model for future research.     

Performance of genetic algorithms in search for water splitting perovskites

We examine the performance of genetic algorithms (GAs) in uncovering solar water light splitters over a space of almost 19,000 perovskite materials. The entire search space was previously calculated using density functional theory to determine solutions that fulfill constraints on stability, band gap, and band edge position. Here, we test over 2500 unique GA implementations in finding these solutions to determine whether GA can avoid the need for brute force search, and thereby enable larger chemical spaces to be screened within a given computational budget. We find that the best GAs tested offer almost a 6 times efficiency gain over random search, and are comparable to the performance of a search based on informed chemical rules. In addition, the GA is almost 10 times as efficient as random search in finding half the solutions within the search space. By employing chemical rules, the performance of the GA can be further improved to approximately 12–17 better than random search. We discuss the effect of population size, selection function, crossover function, mutation rate, fitness function, and elitism on the final result, finding that selection function and elitism are especially important to GA performance. In addition, we determine that parameters that perform well in finding solar water splitters can also be applied to discovering transparent photocorrosion shields. Our results indicate that coupling GAs to high-throughput density functional calculations presents a promising method to rapidly search large chemical spaces for technological materials.     

Emerging specializations,competences and firms' proximity in digital industries: The case of London

In order to overcome the limitations of defining industrial specializations in digital industries through SIC codes, this paper suggests measuring the specializations and competences of these industries on the basis of the degree of digital technologies present in the products and services supplied. Metadata from CrunchBase are employed, as proxies of firms' specializations and competences which are defined as the fields of activity in which firms are involved. Applying a network analysis, these specializations and competences are linked to the recognition of emerging digital technologies and the strongest combinations of products and services. We tested the proposed methodology on London, a leading centre for the digital economy.     

Vulnerability,Resilience and ‘Systemic Interest’: a Connectivity Approach

Networks and Spatial Economics - This paper analyzes the socio-economic dimension of vulnerability and resilience from the viewpoint of connectivity. While no consensus has yet emerged on...