Résumé
Identifying and exploring emerging trends in news is becoming more essential than ever with many changes occurring around the world due to the global health crises. However, most of the recent research has focused mainly on detecting trends in social media, thus, benefiting from social features (e.g. likes and retweets on Twitter) which helped the task as they can be used to measure the engagement and diffusion rate of content. Yet, formal text data, unlike short social media posts, comes with a longer, less restricted writing format, and thus, more challenging. In this paper, we focus our study on emerging trends detection in financial news articles about Microsoft, collected before and during the start of the COVID-19 pandemic (July 2019 to July 2020). We make the dataset accessible and we also propose a strong baseline (Contextual Leap2Trend) for exploring the dynamics of similarities between pairs of keywords based on topic modeling and term frequency. Finally, we evaluate against a gold standard (Google Trends) and present noteworthy real-world scenarios regarding the influence of the pandemic on Microsoft. ©2022 Association for Computational Linguistics.
Résumé
Introduction: Mental health regional differences during pregnancy through the COVID-19 pandemic is understudied. Objective(s): We aimed to quantify the impact of the COVID-19 pandemic on maternal mental health during pregnancy. Method(s): A cohort study with a web-based recruitment strategy and electronic data collection was initiated in 06/2020. Although Canadian women, >18 years were primarily targeted, pregnant women worldwide were eligible. The current analysis includes data on women enrolled 06/2020-11/2020. Self-reported data included mental health measures (Edinburgh Perinatal Depression Scale (EPDS), Generalized Anxiety Disorders (GAD-7)), stress. We compared maternal mental health stratifying on country/continents of residence, and identified determinants of mental health using multivariable regression models. Result(s): Of 2,109 pregnant women recruited, 1,932 were from Canada, 48 the United States (US), 73 Europe, 35 Africa, and 21 Asia/Oceania. Mean depressive symptom scores were lower in Canada (EPDS 8.2, SD 5.2) compared to the US (EPDS 10.5, SD 4.8) and Europe (EPDS 10.4, SD 6.5) (p<0.05), regardless of being infected or not. Maternal anxiety, stress, decreased income and access to health care due to the pandemic were increasing maternal depression. The prevalence of severe anxiety was similar across country/continents. Maternal depression, stress, and earlier recruitment during the pandemic (June/July) were associated with increased maternal anxiety. Conclusion(s): In this first international study on the impact of the COVID-19 pandemic, CONCEPTION has shown significant country/continent-specific variations in depressive symptoms during pregnancy, whereas severe anxiety was similar regardless of place of residence. Strategies are needed to reduce COVID-19's mental health burden in pregnancy.
Résumé
Tracking news stories in documents is a way to deal with the large amount of information that surrounds us everyday, to reduce the noise and to detect emergent topics in news. Since the Covid-19 outbreak, the world has known a new problem: infodemic. News article titles are massively shared on social networks and the analysis of trends and growing topics is complex. Grouping documents in news stories lowers the number of topics to analyse and the information to ingest and/or evaluate. Our study proposes to analyse news tracking with little information provided by titles on social networks. In this paper, we take advantage of datasets of public news article titles to experiment news tracking algorithms on short messages. We evaluate the clustering performance with little amount of data per document. We deal with the document representation (sparse with TF-IDF and dense using Transformers [26]), its impact on the results and why it is key to this type of work. We used a supervised algorithm proposed by Miranda et al. [22] and K-Means to provide evaluations for different use cases. We found that TF-IDF vectors are not always the best ones to group documents, and that algorithms are sensitive to the type of representation. Knowing this, we recommend taking both aspects into account while tracking news stories in short messages. With this paper, we share all the source code and resources we handled. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
Résumé
In the past decade, multiple compounding crises-ecological, racial injustices, 'care crises' and multiple recent crises related to the COVID-19 pandemic-have reinforced the powerful role of critical and social policy researchers to push back against 'fake news', alternative facts', and a post-truth era that denigrates science and evidence-based research. These new realities can pose challenges for social scientists who work within relational, ontological, non-representational, new materialist, performative, decolonising, or ecological 'turns' in social theory and epistemologies. This article's overarching question is: How does one work within non-representational research paradigms while also attempting to hold onto representational, authoritative and convincing versions of truth, evidence, facts and data? Informed by my research on feminist philosopher and epistemologist Lorraine Code's 40-year trajectory of writing about knowledge making and ecological social imaginaries, I navigate these dilemmas by calling on an unexpected ally to family sociology and family policy: The late American environmentalist Rachel Carson. Extending Code's case study of Carson, I argue for an approach that combines (1) ecological relational ontologies, (2) the ethics and politics of knowledge making, (3) crossing social imaginaries of knowledge making and (4) a reconfigured view of knowledge makers as working towards just and cohabitable worlds. © Policy Press 2021.