Notice bibliographique
Résumé
During 2017, Donald J. Trump posted an average of more than seven tweets a day.In the immediately previous years, Trump had employed tweets as a campaign strategy in his run for the U.S. presidency (Enli, 2017;Lee & Lim, 2016;Ott, 2017;Wang, Luo, Niemi, Li,& Hu, 2016), and he continued to tweet at a high rate once in office.Twitter is a relatively new communication technology.Presidents Obama and Trump both tweeted, but Trump tweeted at a much higher rate.As reported in sources described in the Methodology section below, Obama tweeted 321 times in his last 19 months in office, while Trump tweeted 2,602 times in his first year.This research addresses the pleasantness/unpleasantness of the language in Trump's 2017 tweets.It questions whether this pleasantness/unpleasantness is related to tweet characteristics (such as their length in number of words, their inclusion of a hash tag, or their status as a reply) or to information included within tweet text (such as an American flag emoji, an exclamation mark, or a mention of CNN).These variables are employed to predict pleasantness, and pleasantness in turn is employed with the other predictors to identify popular tweets that were retweeted at a high rate. Measuring the Pleasantness/Unpleasantness of TweetsSentiment analysis is an approach that examines the emotion conveyed-in tweets, in this case-by looking at the words in them.Some forms of sentiment analysis depend on lists of words indicative of various emotional states; they evaluate the occurrence of these words in tweets and infer emotionality from occurrence, sometimes in very sophisticated ways (e.g., Mohammad, Kiritchenko, & Zhu, 2013;Saif, Fernandez, He, & Alani, 2016).The approach employed in this research depends instead on the rated emotional connotations of many thousands of words, including some very common ones (Whissell, 2009).Ratings contained in the Dictionary of Affect in Language (Whissell, 2009) 1 were provided by multiple raters.They had been obtained, several years before the current research, in a totally independent setting where participants rated context-free words on their pleasantness.The scale employed in this research is a simple linear transformation of the original scale.The tool has been applied to many different types of data: from Shakespeare (Whissell, 2010a) to President Clinton's communications (Whissell, 2010b). 1 Referred to in the rest of this article as the Dictionary, or the Dictionary of Affect Abstract: President Trump's 2017 tweets (N = 2,602) were studied in terms of their pleasantness (as measured by the Dictionary of Affect; Whissell, 2009), their popularity (defined in terms of retweets and likes), message characteristics (e.g., message length), and message content (e.g., whether the message included mentions of Melania, CNN, Hillary Clinton, or Democrats).Pleasantness and popularity were significantly predicted (p<.001) on the basis of message characteristics and content variables.Table 3 highlights the presence of two types of message: pleasant/unpopular tweets and unpleasant/popular ones.On the basis of their characteristics and contents, the first type of tweet was labeled as celebratory/ congratulatory and the second as antagonistic/accusatory. Antagonistic/accusatory tweets tended to mention Trump's political opponents (e.g., Clinton, Democrats, CNN) and to be posted between 7:00 pm and midnight.Trump's tweets were mildly pleasant in tone, but not as pleasant as tweets posted by President Obama in 2015-2016.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,008 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,000 |
| Méta-épidémiologie (sens large) | 0,000 | 0,000 |
| Bibliométrie | 0,004 | 0,004 |
| Études des sciences et des technologies | 0,002 | 0,000 |
| Communication savante | 0,003 | 0,003 |
| Science ouverte | 0,001 | 0,002 |
| Intégrité de la recherche | 0,001 | 0,002 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,098 | 0,066 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».