Search

Remote sensing (RS) of volcanic gases has become a central tool for studying volcanic activity. For instance, ultraviolet (UV) skylight spectroscopy with grating spectrographs (GS) enables SO2 (and, under favourable conditions, BrO) quantification in volcanic plumes from autonomous platforms at safe distances. These measurements can serve volcanic monitoring and they cover all stages of volcanic activity in long measurement time series, which substantially contributes to the refinement of theories on volcanic degassing. Infrared (IR) remote sensing techniques are able to measure further volcanic gases (e.g., HF, HCl, CO2, CO). However, the employed Fourier transform spectrometers (FTSs) are intrinsically intricate and, due to limited resolving power or light throughput, mostly rely on either lamps, direct sun, or hot lava as light source, usually limiting measurements to individual field campaigns. We show that many limitations of grating spectrographs and Fourier transform spectrometer measurements can be overcome by Fabry-Perot interferometer (FPI) based spectrograph implementations. Compared to grating spectrographs and Fourier transform spectrometers, Fabry-Perot interferometer spectrographs reach a 1-3 orders of magnitude higher spectral resolution and superior light throughput with compact and stable set-ups. This leads to 1) enhanced sensitivity and selectivity of the spectral trace gas detection, 2) enables the measurement of so far undetected volcanic plume constituents [e.g., hydroxyl (OH) or sulfanyl (SH)], and 3) extends the range of gases that can be measured continuously using the sky as light source. Here, we present measurements with a shoe-box-size Fabry-Perot interferometer spectrograph (resolving power of ca. 150000), performed in the crater of Nyiragongo volcano. By analysing the light of a ultraviolet light emitting diode that is sent through the hot gas emission of an active lava flow, we reach an OH detection limit of about 20 ppb, which is orders of magnitude lower than the mixing ratios predicted by high-temperature chemical models. Furthermore, we introduce example calculations that demonstrate the feasibility of skylight-based remote sensing of HF and HCl in the short-wave infrared with Fabry-Perot interferometer spectrographs, which opens the path to continuous monitoring and data acquisition during all stages of volcanic activity. This is only one among many further potential applications of remote sensing of volcanic gases with high spectral resolution.

The massive diffusion of social media has produced disintermediation and it has changed the way in which users inform themselves and participate in public debate. On the other hand, users show a tendency to interact with information that adheres to personal choices and previous opinions. This propensity is “exploited” by algorithms that manage and sort communication on social media, increasingly producing a polarized audience. The essay shows the outcomes of these dynamics by focusing on the communications structure related to the pandemic. In this sense, social media constitute a risk because they convey out-of-control access to unreliable information and because they can result in the absence of correct information, if not in a problem for democracy. On the technical level, at least three complex reticular structures are identified that evolve following different directions, emerged from three sequentially monitored lemmas (no-mask, covid-19, greenpass). Further considerations concern the increasingly dense interweaving between online communication and the genesis of protest movements with evanescent, single-issue structures.

In early January 2020, after China reported the first cases of the new coronavirus (SARS-CoV-2) in the city of Wuhan, unreliable and not fully accurate information has started spreading faster than the virus itself. Alongside this pandemic, people have experienced a parallel infodemic, i.e., an overabundance of information, some of which is misleading or even harmful, which has widely spread around the globe. Although social media are increasingly being used as the information source, web search engines, such as Google or Yahoo!, still represent a powerful and trustworthy resource for finding information on the Web. This is due to their capability to capture the largest amount of information, helping users quickly identify the most relevant, useful, although not always the most reliable, results for their search queries. This study aims to detect potential misleading and fake contents by capturing and analysing textual information, which flow through search engines. By using a real-world dataset associated with recent COVID-19 pandemic, we first apply re-sampling techniques for class imbalance, and then we use existing machine learning algorithms for classification of not reliable news. By extracting lexical and host-based features of associated uniform resource locators (URLs) for news articles, we show that the proposed methods, so common in phishing and malicious URL detection, can improve the efficiency and performance of classifiers. Based on these findings, we suggest that the use of both textual and URL features can improve the effectiveness of fake news detection methods.

This paper proposes the use of data mining techniques for the analysis of criminal sentences. In particular, we are interested in the criminal sentences emitted for mafia-related crimes (ex art. 51, co 3 bis c.p.p.) by the four districts of the Appeal Courts of Sicily from 2000 up to 2006, and deemed irrevocable for at least one defendant.This research is particularly important for three reasons. (1) There is currently a lack of comparative analysis of the four districts of the Appeal Courts of Sicily (2) there is currently no collection of data describing the institutional answer to the mafia phenomenon; (3) Sicilian jurisprudence leads the Italian jurisprudence for this kind of phenomenon. The main objective is to extract from the analysis of the text of the sentences a description of the development of the sentences, as well as the socio-economic scope in which the events relative to the sentences happened.The research has shown that sentences can provide a lot of information; in particular if one takes into account the juridical statistics, their insufficient reliability and their lack of informative character in relation to the recipients of the data, and the difficulty in collecting and comparing the data, that is caused by the speed of the legislative changes and by the classification criteria.Thanks to the close collaboration between social scientists and computer scientists, we have developed an ad hoc reading algorithm, that analyzes the internal structure of the sentences, and that can be generalized to other areas of the jurisprudence, as well as to other temporal and geographical scopes.

In this paper we discuss a data mining application to profile online users on a news website. Such profiles are then used to target users with personalized content in order to increase the overall number of pageviews generated. The additional pageviews provide extra revenue for the online editor. While the proposed system satisfies the online editor to better monetize their content, in our study, we also provide a sociological interpretation of the produced knowledge base. We consider this to be a quite novel approach where scientists from two very different disciplines (i.e., computer science and social science) combine their effort in order to provide a common framework for interpreting and utilizing the information extracted. Both disciplines benefit from a combined approach. From one side the computer scientists can “tune” their data mining algorithms in order to support the different sociological view of the data. From the other side, the social scientists can benefit from the additional information provided by the data mining tool as they provide a new insight for their data interpretation.