Cognitive Networks Extract Insights on COVID-19 Vaccines from English and Italian Popular Tweets: Anticipation, Logistics, Conspiracy and Loss of Trust

Stella, Massimo; Vitevitch, Michael S.; Botta, Federico

doi:10.3390/bdcc6020052

Cited by 12 publications

(12 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…social media posts [5,14,19], movie transcripts [37] and books [25,38,39]. The structure of cognitive networks was shown to highlight key cognitive patterns like writing styles [39], alterations to cognition due to psychedelic drugs [36], and, relevantly for this work, it could reconstruct the semantic and emotional content of online perceptions towards to gender gap in science [15] and COVID-19 vaccines [5,40]. In text analysis, cognitive networks are advantageous for reconstructing the semantic frame of a target concept, i.e.…”

Section: Relevant Literaturementioning

confidence: 82%

Natural Language Processing and Cognitive Networks Identify UK Insurers’ Trends in Investor Day Transcripts

Stefan¹,

Stella²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

The ability to identify and rank risk is essential for efficient and effective supervision of financial service firms, such as banks and insurers. Risk ranking ensures limited resources are allocated where they are most needed. Today, automatic risk identification within insurance supervision primarily relies on quantitative metrics based on numerical data (e.g. returns). The purpose of this work is to assess whether Natural Language Processing (NLP) and cognitive networks can achieve similar automated risk ranking and identification by analysing textual data, i.e. NIDT=829 investor transcripts from Bloomberg. To this aim, this work explores and tunes 3 NLP techniques: (1) keyword extraction enhanced by cognitive network analysis; (2) valence/sentiment analysis; and (3) topic modelling. Results highlight that keyword analysis, enriched by term frequency-inverse document frequency scores and semantic framing through cognitive networks, could detect events of relevance for the insurance system like cyber-attacks or the COVID-19 pandemic. Cognitive networks were found to highlight events that related to specific financial transitions: The semantic frame of "climate" grew in size by +538% between 2018 and 2020 and outlined an increased awareness that agents and insurers expressed towards climate change. A lexicon-based sentiment analysis achieved a Pearson’s correlation of ρ=0.16 (p&lt;0.001,N=829) between sentiment levels and daily share prices. Although relatively weak, this finding indicates that insurance jargon is insightful to support risk supervision. Topic modelling is considered less amenable to support supervision, because of a lack of results’ stability and an intrinsic difficulty to interpret risk patterns. We discuss how these automatic methods could complement existing supervisory tools in automated risk ranking.

show abstract

Section: Relevant Literaturementioning

confidence: 82%

Natural Language Processing and Cognitive Networks Identify UK Insurers’ Trends in Investor Day Transcripts

Stefan¹,

Stella²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Regarding the most recent works by the authors who receive the most citations, some of them, including Rickheit [ 107 ], Herrmann [ 122 ], and Brysbaert [ 91 ], address general topics such as the history of psycholinguistics [ 107 ], language use [ 122 ], reading [ 123 ], and individual differences [ 124 , 125 ]. Other authors, such as Levelt [ 3 ], address more specific topics, including visual word recognition [ 126 ], semantic frames [ 127 ], bilingualism [ 128 ], and grammar learning [ 129 ].…”

Section: Discussionmentioning

confidence: 99%

Psycholinguistics: Analysis of Knowledge Domains on Children’s Language Acquisition, Production, Comprehension, and Dissolution

et al. 2022

View full text Add to dashboard Cite

This paper utilised bibliometric and scientometric indicators to assess the current state of research in psycholinguistics. A total of 32,586 documents in psycholinguistics were included from Scopus, WOS, and Lens between 1946 and 2022. The collected data were analysed using CiteSpace 5.8.R3 and VOSviewer 1.6.18. The results included tabulation, visualisation, and mapping for the past, present, and future directions of the field of psycholinguistics. We identified key authors, works, journals, and concepts in the existing evidence concerning (children’s) language acquisition, production, comprehension, and dissolution. The study contributes to the systematic study of existing scholarship in the field of psycholinguistics by documenting the progress of the field and informing relevant researchers about the current state of the field of psycholinguistics. Having grouped the 32,586 documents in psycholinguistics, 12 clusters were identified. These include (1) examining individual difference in affective norm and familiarity account; (2) examining refractory effect in the role of Broca’s area in sentence processing; (3) using eye movement to study bilingual language control and familiarity account; (4) exploring familiarity account through relative clauses; (5) the study of formulaic language and language persistence; (6) examining affective norm and sub-lexical effect in Spanish words; (7) examining lexical persistence in multiplex lexical networks; (8) the study of persistence through cortical dynamics; (9) the study of context effect in language learning and language processing; (10) the study of neurophysiological correlates in semantic context integration; (11) examining persistence as an acquisition norm through naming latencies; and (12) following a cross-linguistic perspective to study aphasic speakers.

show abstract

“…Estudios algorítmicos de la red social Twitter, como el de Zheng et al (2021) y Doha et al (2022, exponen cómo los tuits con sentimientos y emociones negativas durante la pandemia por COVID-19 fueron más difundidos que los positivos, mientras que estudios en esta misma red social en inglés y en italiano han mostrado una gran polarización especialmente contra la vacuna AstraZeneca-Oxford y una mayor tristeza (Stella et al, 2022), al igual que en India (Kumar, 2022) o Alemania (Fieselmann et al, 2022), unidos a la presencia de una gran cantidad de desinformación alrededor de las vacunas (Ng et al, 2022).…”

Section: Vacunas Redes Sociales Y Opinión Públicaunclassified

“…Esto vendría a sugerir la posibilidad de la existencia de grupos negacionistas y/o anti-vacunas organizados a través de pequeñas cuentas anónimas que buscan sembrar la duda y la polarización, aprovechando principalmente los casos aparecidos en medios con la vacuna AstraZeneca-Oxford y que se intenta extender a otras vacunas, como la de Pfizer-BioNTech o la de Janssen. Este ataque a las vacunas, especialmente la de AstraZeneca-Oxford ha sido corroborada igualmente en otros países (Stella et al, 2022). Esta sería la técnica del llamado efecto astroturfing, campañas donde "la difusión de opiniones engañosas por impostores [humanos o no] que se hacen pasar por personas autónomas en Internet con la intención de impulsar una agenda específica" (Zhang et al, 2013) y que se enlaza a la propagación de desinformación alrededor de la vacunación como expone Ng et al (2022).…”

Section: Conclusionesunclassified

Vacunas anticovid y trombosis: el miedo en las redes sociales

García

Garcinuño

Cano

et al. 2023

RCyS

View full text Add to dashboard Cite

Fundamentos: El presente artículo tiene como objetivo analizar la relación existente en la red social Twitter en español entre trombos y vacunas anticovid. Métodos: Se utilizaron técnicas de machine learning y análisis estadístico masivo para la determinación de redes sociales formadas, sus relaciones, análisis de sentimientos y emociones y discursos principales por minería de texto durante cuatro meses. Para ello se recogieron 915.825 mensajes que hablaran de trombosis a lo largo de los cuatro primeros meses del año 2021. Resultados: Se encontró un aumento muy destacado de mensajes alrededor desde mediados de marzo e incrementándose fuertemente a principios de abril, coincidiendo con los casos detectados y suspensiones temporales de ciertas vacunas en Estados Unidos y Europa. El principal de mensajes procede de España, aunque México centra el debate en Latinoamérica. Conclusiones: Se pudieron determinar un aumento de mensajes de alta carga emocional, principalmente negativa, así como mensajes desinformativos y conspirativos especialmente desde unos grupos sin referentes significativos. Los difusores de noticias de desinformación sobre vacunas son enviados por pequeñas cuentas nano-influencers o cuentas falsas utilizando posibles técnicas astroturfing.

show abstract

Cognitive Networks Extract Insights on COVID-19 Vaccines from English and Italian Popular Tweets: Anticipation, Logistics, Conspiracy and Loss of Trust

Cited by 12 publications

References 44 publications

Natural Language Processing and Cognitive Networks Identify UK Insurers’ Trends in Investor Day Transcripts

Natural Language Processing and Cognitive Networks Identify UK Insurers’ Trends in Investor Day Transcripts

Psycholinguistics: Analysis of Knowledge Domains on Children’s Language Acquisition, Production, Comprehension, and Dissolution

Vacunas anticovid y trombosis: el miedo en las redes sociales

Contact Info

Product

Resources

About