arXiv subjects

Genoveva Vargas-Solar

Publications and source records attributed to Genoveva Vargas-Solar.

24 records · Page 2Linked to original sources

arXiv2021

LACLICHEV: Exploring the History of Climate Change in Latin America within Newspapers Digital Collections

This paper introduces LACLICHEV (Latin American Climate Change Evolution platform ), a data collections exploration environment for exploring historical newspapers searching for articles reporting meteorological events. LACLICHEV is based on data collections' exploration techniques combined with information retrieval, data analytics, and geographic querying and visualization. This environment provides tools for curating, exploring and analyzing historical newspapers articles, their description and location, and the vocabularies used for referring to meteorological events. The objective being to understand the content of newspapers and identifying possible patterns and models that can build a view of the history of climate change in the Latin American region.

cs.DL↗

arXiv2021

A Geo-Gender Study of Indexed Computer Science Research Publications

This paper presents a study that analyzes and gives quantitative means for measuring the gender gap in computing research publications. The data set built for this study is a geo-gender tagged authorship database named authorships that integrates data from computing journals indexed in the Journal Citation Reports (JCR) and the Microsoft Academic Graph (MAG). We propose a gender gap index to analyze female and male authors' participation gap in JCR publications in Computer Science. Tagging publications with this index, we can classify papers according to the degree of participation of both women and men in different domains. Given that working contexts vary for female scientists depending on the country, our study groups analytics results according to the country of authors affiliation institutions. The paper details the method used to obtain, clean and validate the data, and then it states the hypothesis adopted for defining our index and classifications. Our study results have led to enlightening conclusions concerning various aspects of female authorship's geographical distribution in computing JCR publications.

cs.DL↗

arXiv2021

Looking for COVID-19 misinformation in multilingual social media texts

This paper presents the Multilingual COVID-19 Analysis Method (CMTA) for detecting and observing the spread of misinformation about this disease within texts. CMTA proposes a data science (DS) pipeline that applies machine learning models for processing, classifying (Dense-CNN) and analyzing (MBERT) multilingual (micro)-texts. DS pipeline data preparation tasks extract features from multilingual textual data and categorize it into specific information classes (i.e., 'false', 'partly false', 'misleading'). The CMTA pipeline has been experimented with multilingual micro-texts (tweets), showing misinformation spread across different languages. To assess the performance of CMTA and put it in perspective, we performed a comparative analysis of CMTA with eight monolingual models used for detecting misinformation. The comparison shows that CMTA has surpassed various monolingual models and suggests that it can be used as a general method for detecting misinformation in multilingual micro-texts. CMTA experimental results show misinformation trends about COVID-19 in different languages during the first pandemic months.

cs.CL↗

arXiv2021

Putting Data Science Pipelines on the Edge

This paper proposes a composable "Just in Time Architecture" for Data Science (DS) Pipelines named JITA-4DS and associated resource management techniques for configuring disaggregated data centers (DCs). DCs under our approach are composable based on vertical integration of the application, middleware/operating system, and hardware layers customized dynamically to meet application Service Level Objectives (SLO - application-aware management). Thereby, pipelines utilize a set of flexible building blocks that can be dynamically and automatically assembled and re-assembled to meet the dynamic changes in the workload's SLOs. To assess disaggregated DC's, we study how to model and validate their performance in large-scale settings.

cs.DB↗

arXiv2020

From Data Harvesting to Querying for Making Urban Territories Smart

This chapter provides a summarized, critical and analytical point of view of the data-centric solutions that are currently applied for addressing urban problems in cities. These solutions lead to the use of urban computing techniques to address their daily life issues. Data-centric solutions have become popular due to the emergence of data science. The chapter describes and discusses the type of urban challenges and how data science in urban computing can face them. Current solutions address a spectrum that goes from data harvesting techniques to decision making support. Finally, the chapter also puts in perspective families of strategies developed in the state of the art for addressing urban problems and exhibits guidelines that can lead to a methodological understanding of these strategies.

cs.DB↗

arXiv2019

Analyzing digital politics: Challenges and experiments in a dual perspective

Social networks have become in the last decade central to political life. However, to those interested in analysing the communication strategies of parties and candidates at election time, the introduction of the Internet into the political sphere has proved a mixed blessing. Indeed, while retrieving, consulting, and archiving original documents pertaining to a specific campaign have become easier, faster, and achievable on a larger scale, thus opening up a promising El Dorado for research in this area, studying online campaigns has also inevitably introduced new technical, methodological and legal challenges which have turned out to be increasingly complex for academics in the humanities and social sciences to solve on their own.This paper therefore proposes to provide feedback on experience and experimental validation from a multidisciplinary project called POLIWEB devoted to the comparative analysis of political campaigns on social media in the run up to the 2014 elections to the European Parliament in France and in the United Kingdom. Together with observations from a humanities' perspective on issues related to such a project, this paper also presents experimental results concerning three of the data collection life cycle phases: collection, cleaning, and storage. The outcome is a data collection ready to be analysed for various purposes meant to address the political science topic under consideration.

cs.SI↗