ArticlePDF Available

Tweet topics and sentiments relating to distance learning among Italian Twitter users

Abstract and Figures

The outbreak of COVID-19 forced a dramatic shift in education, from in-person learning to an increased use of distance learning over the past 2 years. Opinions and sentiments regarding this switch from traditional to remote classes can be tracked in real time in microblog messages promptly shared by Twitter users, who constitute a large and ever-increasing number of individuals today. Given this framework, the present study aims to investigate sentiments and topics related to distance learning in Italy from March 2020 to November 2021. A two-step sentiment analysis was performed using the VADER model and the syuzhet package to understand the overall sentiments and emotions. A dynamic latent Dirichlet allocation model (DLDA) was built to identify commonly discussed topics in tweets and their evolution over time. The results show a modest majority of negative opinions, which shifted over time until the trend reversed. Among the eight emotions of the syuzhet package, ‘trust’ was the most positive emotion observed in the tweets, while ‘fear’ and ‘sadness’ were the top negative emotions. Our analysis also identified three topics: (1) requests for support measures for distance learning, (2) concerns about distance learning and its application, and (3) anxiety about the government decrees introducing the red zones and the corresponding restrictions. People’s attitudes changed over time. The concerns about distance learning and its future applications (topic 2) gained importance in the latter stages of 2021, while the first and third topics, which were ranked highly at first, started a steep descent in the last part of the period. The results indicate that even if current distance learning ends, the Italian people are concerned that any new emergency will bring distance learning back into use again.
This content is subject to copyright. Terms and conditions apply.
1
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports
Tweet topics and sentiments
relating to distance learning
among Italian Twitter users
Luisa Stracqualursi* & Patrizia Agati
The outbreak of COVID-19 forced a dramatic shift in education, from in-person learning to an
increased use of distance learning over the past 2 years. Opinions and sentiments regarding this switch
from traditional to remote classes can be tracked in real time in microblog messages promptly shared
by Twitter users, who constitute a large and ever-increasing number of individuals today. Given this
framework, the present study aims to investigate sentiments and topics related to distance learning
in Italy from March 2020 to November 2021. A two-step sentiment analysis was performed using
the VADER model and the syuzhet package to understand the overall sentiments and emotions. A
dynamic latent Dirichlet allocation model (DLDA) was built to identify commonly discussed topics
in tweets and their evolution over time. The results show a modest majority of negative opinions,
which shifted over time until the trend reversed. Among the eight emotions of the syuzhet package,
‘trust’ was the most positive emotion observed in the tweets, while ‘fear’ and ‘sadness’ were the top
negative emotions. Our analysis also identied three topics: (1) requests for support measures for
distance learning, (2) concerns about distance learning and its application, and (3) anxiety about the
government decrees introducing the red zones and the corresponding restrictions. People’s attitudes
changed over time. The concerns about distance learning and its future applications (topic 2) gained
importance in the latter stages of 2021, while the rst and third topics, which were ranked highly at
rst, started a steep descent in the last part of the period. The results indicate that even if current
distance learning ends, the Italian people are concerned that any new emergency will bring distance
learning back into use again.
e COVID-19 pandemic has greatly aected life worldwide. One of the most remarkable eects was the enforce-
ment of social distancing to reduce the spread of the disease. In March 20201, Italy implemented social-distancing
measures by enforcing distance learning at all educational stages and online assessments to help continue stu-
dents’ education2. ese measures became known as ‘emergency distance learning’ and introduced new experi-
ences and challenges for students, parents, and teachers. In the subsequent months, distance learning gradually
moved to ‘integrated digital learning3, which combined remote (virtual classroom) and in-person (traditional
classroom) instruction. Unfortunately, this integration was very slow: the reopening of schools has been limited
to some Italian regions and has oen been only temporary. As post-outbreak SARS-CoV-2 infections increased,
many regions suddenly returned to distance learning for either some grades of school or for all, as happened in
Italy’s ‘red zones’.
Social media has been a major and rich data source for research in many domains due to its 3.8 billion active
users4 across the globe. For instance, researchers analyze user comments extracted from social media platforms
(such as Facebook5, Twitter5, and Instagram6) to uncover insights about social issues such as health, politics and
business. Among these platforms, Twitter stands out as one of the most immediate; tweets ow nonstop on the
bulletin boards of users incessantly. Twitter allows users to express and spread opinions, thoughts and emotions as
concisely and quickly as possible. erefore, researchers have oen preferred to analyze user comments on Twitter
to immediately uncover insights about social issues during the coronavirus pandemic (e.g., conspiracy theories7,
why people oppose wearing a mask8, experiences in health care9, and vaccinations10) or distance learning1113.
e text content of a tweet is a short microblog message containing at most 280 characters; this feature makes
tweets particularly suitable for natural language processing (NLP) techniques, which are widely used to extract
insights from unstructured texts. Distance learning was much debated during the pandemic. On the other hand,
we chose Twitter for its immediacy in capturing and spreading people’s opinions and emotions on any topic, as
well as for its ability to provide plentiful data, even in a short amount of time. Moreover, the people who have
OPEN
Department of Statistics, University of Bologna, 40126 Bologna, Italy. *email: luisa.stracqualursi@unibo.it
Content courtesy of Springer Nature, terms of use apply. Rights reserved
2
Vol:.(1234567890)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
more directly experienced distance learning are students, parents, and teachers, that is, people who, by age, make
up approximately 83% of Twitter users46.
is study aims to explore sentiments and major topics about distance learning in Italy and their evolution
over time by using NLP techniques to analyze tweets from Italian Twitter users. Findings from this study could
help the Ministry of Education visualize how people are coping with distance learning, thus improving distance
learning support and making the experience more eective in the future.
Unlike traditional methods, which are expensive and time-consuming even for small samples, NLP techniques
use big data and social media and are very economic, fast, and immediate. A well-known drawback of these
methods, however, is that they do not allow us to consider social variables (e.g., age, gender, marital status, mode
of working) related to the emotions revealed by the model.
In the literature, COVID-19 has been associated with psychological distress, depression, anxiety, and fear1416.
Other research highlights a signicant level of traumatic stress in women more than in men17. Moreover, pregnant
women during lockdowns suered the most from anxiety and depression18.
Regarding age, the research highlights that older people suered the most from negative eects such as fear
and loneliness19,20. Younger individuals had fewer negative emotions because they saw COVID-19 as a less risky
disease for them21, although they did report anxiety and depression due to the social restrictions imposed21.
Finally, regarding marital status, Rania and Coppola22 show how single, divorced and separated individuals
were the most aected by loneliness and demonstrated a higher level of mental illness compared to married
individuals. In addition, dierences also emerged regarding work during COVID-19. ose who continued to
work without changes reported a lower level of mental health than those who switched to working remotely.
Methodology
The data. Twitter was chosen as the data source. It is one of the worlds major social media platforms, with
199 million active users in April 20214, and it is also a common source of text for sentiment analyses2325.
To collect distance learning-related tweets, we used TrackMyHashtag https:// www. track myhas htag. com/, a
tracking tool to monitor hashtags in real time. Unlike Twitter API, which does not provide tweets older than
three weeks, TrackMyHashtag also provides historical data and lters selections by language and geolocation.
For our study, we chose the Italian words for ‘distance learning’ as the search term and selected March 3,
2020 through November 23, 2021 as the period of interest. Finally, we chose Italian tweets only. A total of 25,100
tweets were collected for this study.
Data preprocessing. To clean the data and prepare it for sentiment analysis, we applied the following pre-
processing steps using NLP techniques implemented with Python:
1. removed mentions, URLs, and hashtags,
2. replaced HTML characters with Unicode equivalent (such as replacing ‘&’ with ‘&’),
3. removed HTML tags (such as
<div >
,
<p>
, etc.),
4. removed unnecessary line breaks,
5. removed special characters and punctuation,
6. removed words that are numbers,
7. converted the Italian tweets’ text into English using the ‘googletrans’ tool.
In the second part an higher quality dataset is required for the topic model. e duplicate tweets were
removed, and only the unique tweets were retained. Apart from the general data-cleaning methods, tokenization
and lemmatization could enable the model to achieve better performance. e dierent forms of a word cause
misclassication for models. Consequently, the WorldNet library of NLTK26 was used to accomplish lemmatiza-
tion. e stemming algorithms that aggressively reduce words to a common base even if these words actually
have dierent meanings are not considered here. Finally, we lowercased all of the text to ensure that every word
appeared in a consistent format and pruned the vocabulary, removing stop words and terms unrelated to the
topic, such as ‘as’, ‘from, and ‘would’.
Sentiment and emotion analysis. Between the major algorithms to be used for text mining and spe-
cically for sentiment analysis, we applied the Valence Aware Dictionary for Sentiment Reasoning (VADER)
proposed by Hutto etal.27 to determine the polarity and intensity of the tweets. VADER is a sentiment lexicon
and rule-based sentiment analysis tool obtained through the wisdom of the crowd approach. rough extensive
human work, this tool enables the sentiment analysis of social media to be completed quickly and has a very high
accuracy similar to that of human beings. We used VADER to obtain sentiment scores for a tweet’s preprocessed
text data. At the same time, according to the classication method recommended by its authors, we mapped the
emotional score into three categories: positive, negative, and neutral (Fig.1 step1).
en, to discover the emotions underlying categories, we applied the nrc28 algorithm, which is one of the
methods included in the R library package syuzhet29 for emotion analysis. In particular, the nrc algorithm applies
an emotion dictionary to score each tweet based on two sentiments (positive or negative) and eight emotions
(anger, fear, anticipation, trust, surprise, sadness, joy, and disgust). Emotional recognition aims to identify the
emotions that a tweet carries. If a tweet was associated with a particular emotion or sentiment, it scores points
that reect the degree of valence with respect to that category. Otherwise, it would have no score for that cat-
egory. erefore, if a tweet contains two words listed in the list of words for the ‘joy’ emotion, the score for that
sentence in the joy category will be 2.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
3
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
When using the nrc lexicon, rather than receiving the algebraic score due to positive and negative words,
each tweet obtains a score for each emotion category. However, this algorithm fails to properly account for nega-
tors. Additionally, it adopts the bag-of-words approach, where the sentiment is based on the individual words
occurring in the text, neglecting the role of syntax and grammar. erefore, the VADER and nrc methods are
not comparable in terms of the number of tweets and polarity categories. Hence, the idea is to use VADER for
sentiment analysis and subsequently to apply nrc only to discover positive and negative emotions. e ow chart
in Fig.1 represents the two-step sentiment analysis. VADER’s neutral tweets are very useful in the classication
but not interesting for the emotions analysis; therefore, we focused on tweets with positive and negative senti-
ments. VADER’s performance in the eld of social media text is excellent. Based on its complete rules, VADER
can carry out a sentiment analysis on various lexical features: punctuation, capitalization, degree modiers, the
contrastive conjunction ‘but’, and negation ipping tri-grams.
The topic model. e topic model is an unsupervised machine learning method; that is, it is a text mining
procedure with which the topics or themes of documents can be identied from a large document corpus30. e
latent Dirichlet allocation (LDA) model is one of the most popular topic modeling methods; it is a probabilistic
model for expressing a corpus based on a three-level hierarchical Bayesian model. e basic idea of LDA is that
each document has a topic, and a topic can be dened as a word distribution31. Particularly in LDA models, the
generation of documents within a corpus follows the following process:
1. A mixture of k topics,
θ
, is sampled from a Dirichlet prior, which is parameterized by
α
;
2. A topic
zn
is sampled from the multinomial distribution,
p|α)
that is the document topic distribution
which models
p(zn=i|θ)
;
3. Fixed the number of topics
, the distribution of words for k topics is denoted by
φ
,which is also
a multinomial distribution whose hyper-parameter
β
follows the Dirichlet distribution;
4. Given the topic
zn
, a word,
wn
, is then sampled via the multinomial distribution
p(w|zn;β)
.
Overall, the probability of a document (or tweet, in our case) “
w
” containing words can be described as:
Finally, the probability of the corpus of M documents
D={w1,...,wM}
can be expressed as the product of
the marginal probabilities of each single document
Dm
, as shown in (2).
(1)
p
(w)=θ
p|α)
N
n=1
k
zn=1
p(wn|zn;β)p(zn|θ)
d
θ
Figure1. Steps ofsentiment and emotion analysis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
4
Vol:.(1234567890)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
In our analysis that includes tweets over a 2-year period, we nd that the tweet content is changeable over
time, and therefore, the topic content is not a static corpus. eDynamic LDA model (DLDA) is adopted and
used on topics aggregated in time epochs, and a state-space model handles transitions of the topics from one
epoch to another. A Gaussian probabilistic model to obtain the posterior probabilities on the evolving topics
along the timeline is added as an additional dimension.
Figure2 shows a graphical representation of the dynamic topic model (DTM)32. As a part of the probabilistic
topic model class, the dynamic model can explain how various tweet themes evolve. e tweet dataset corpus
used here (March 3, 2020-November 23, 2021) contains 630 days, which is exactly seven quarters of a year. e
dynamic topic model is accordingly applied to seven time steps corresponding to the seven trimesters of the
dataset. ese time slices are put into the model provided by gensim33.
An essential challenge in DLDA (as LDA) is to determine an appropriate number of topics. Roder etal. pro-
posed coherence scores to evaluate the quality of each topic model. Particularly, topic coherence is the measure
used to evaluate the coherence between topics inferred by a model. As coherence measures, we used
Cv
and
Cumass
. e rst is a measure based on a sliding window that uses normalized pointwise mutual information
(NPMI) and cosine similarity. Instead,
Cumass
is based on document co-occurrence counts, a one-preceding
segmentation, and a logarithmic conditional probability as conrmation measure. ese values aim to emulate
the relative score that a human is likely to assign to a topic and indicate how much the topic words ‘make sense.
ese scores infer cohesiveness between ‘top’ words within a given topic. Also considered is the distribution
on the primer component analysis (PCA), which can visualize the topic models in a word spatial distribution
with two dimensions. A uniform distribution is preferred, which gives a high degree of independence to each
topic. e judgment for a good model is a higher coherence and an average distribution on the primer analysis
displayed by the pyLDAvis34.
Results
Sentiment analysis. e ndings show that the number of tweets has increased since the beginning of
distance learning (Fig.3). Clearly, visible in the graph, there is a signicant negative sentiment peak on April 22,
2021, due to the Italian government’s ‘reopening decree’ (DL 2021.4.22 no. 52); it xed reopenings of schools and
commercial activities in gradual terms, depending on the degree of epidemic risk in the dierent areas.
Moreover, it is worth noting that the peaks of tweets with a positive sentiment began during the 2021–2022
school year. e highest positive peak was recorded on November 15, triggered by the Italian tax-labor decree
dra. Much hyped by the media, it provided for the renewal of extraordinary leave for parents with children
involved in distance learning. e output of the VADER model, which is the rst step of our sentiment analysis,
shows a modest majority of negative tweets: 8843 negative, 8077 neutral and 8180 positive (35.2%, 32.2% and
(2)
p(D)=
M
m=1θ
pm|α)
Nm
n=1
k
zn=1
p(wm,n|zm,n;β)p(zm,n|θm)
dθ
m
Figure2. Dynamic topic model (for three time slices). A set of topics in the dataset isevolved from the set of
the previous slice. e model for each time slice corresponds to the original LDA process. Additionally, each
topic’s parameters evolve over time.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
5
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
32.6%, respectively). e analysis carried out at the regional level was performed only on 9534 tweets that had
a regional geolocation. Figure4, shows the average sentiment scores of the Italian regions: the sentiment score
is neutral (between − 0.05 and
+
0.05, see Fig.4) for all regions except for Umbria (
+
0.10), Sardinia (
+
0.07)
and Veneto (− 0.06), which slightly exceed the neutrality thresholds. Indeed, there are no major dierences in
school systems in Italy from region to region. Furthermore, the result is consistent with the attening due to the
use of the average of the scores.
e second step of the analysis focuses on searching emotions in nonneutral tweets. Among the eight basic
emotions, ‘trust’ was the prominent positive emotion observed in the tweets, while ‘fear’, ‘sadness’ and ‘anger’
were the top negative emotions (Fig.5). ese results need to be interpreted in light of recent literature on psy-
chological dimensions of the COVID-19 pandemic. e dimension of fear includes the fear of being infected
or infecting others, the risk of death, the loss of loved ones, and not receiving adequate care3538. Several studies
performed during the pandemic found that there is an association between fear and depression14,15,3941. Sadness
is considered by numerous authors to be a core symptom of depression42. e dimensions of anger related to the
pandemic include anger at the government and conspiracy mentalities but also anger at those who fail to comply
with government hygiene measures to contain the virus43.
Figure3. Timeline showing sentiment of tweets about distance learning.
Figure4. Average Sentiment Score in Italy by region.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6
Vol:.(1234567890)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
The topic model. To explore what the user is concerned about on Twitter with reference to distance learn-
ing, we applied LDA to our clean corpus. For a better representation of the entire content, it is necessary to nd
an appropriate topic number. By using topic numbers k ranging from 2 to 10, we initialized the LDA models and
calculated the model coherence. We mainly used
Cv
coherence and
Cumass
coherence as a secondary reference.
According to Fig.6, the coherence score peaked at 3, 4, and 7 topics (6 was not considered because
Cumass
did not
conrm good coherence for this topic). e choice of 4 or 7 topic numbers would lead to a nonuniform distribu-
tion on primer component analysis (PCA), which means that there is not a high degree of independence for each
topic. erefore, we chose 3 as the topic number: the model has no intersections among topics, summarizes the
whole word space well, and the topics remain relatively independent (Fig.7).
In our analysis, we nd that the tweet content changes over time, and therefore, aer initializing through the
LDA model, its dynamic version (DLDA) is used. Our tweets dataset corpus contains 630 days, which makes
exactly seven quarters of a year. e DLDA is accordingly applied to seven time steps corresponding to the seven
trimesters of the dataset. e model output (Fig.8 identied the following three topics:
Topic 1: Digital support
Topic 2: Distance learning concerns
Topic 3: Restriction zones.
e rst theme includes words, such as ‘digital,’ ‘family’ and ‘support’, meaning that people need support in
distance learning. e second topic includes the words ‘work,’ ‘student,’ and ‘lesson. Based on this, we inferred
that most people complain about social issues and personal problems that are dicult to management due to
distance learning. Additionally, several words, such as ‘red,’ ‘zone,’ and ‘ordinance,’ are mentioned in the third
topic. is indicates that a further source of anxiety for the Italians was the government decrees introducing the
red zones and the corresponding restrictions.
Figure5. Emotion analysis of non-neutral tweets performed by syuzhet.
(a) Cv(b) Cumass
Figure6. Coherence values.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
7
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
e dynamic topic model shows that the people’s concerns changed over time (Fig.8). In the topic related
to ‘digital support’, the relevance of words such as ‘family’ and ‘support’ remained stable, while the importance
of the term ‘diculty’ decreased in the later stages of the period. erefore, concerns about support in distance
learning were quite stable over time, while diculties gradually declined.
In the topic related to ‘distance learning concerns’, the importance of words such as ‘school’ and ‘work
remained stable, while the word ‘home’ decreased in importance as time passed until it vanished. ese results
indicate that concerns about distance learning were stable, but the diculty of staying home was no longer one of
them. Last, in the topic related to ‘restriction zones’, the emphasis on words such as ‘covid’ and ‘region’ remained
quite stable, while the term ‘ordinance’ decreased over time. e word ‘zone, which ranked low at rst, started
to climb in the middle of the period and went down again. e main nding indicates an increase in concerns
about restricted zones, following the Italian government decrees establishing the so-called ‘red zones’, i.e., areas
with a high risk of coronavirus infection. e pie charts in Fig.9 show the dynamic volume of each topic in
three periods: March–May 2020, December–February 2021, and September–November 2021. It is worth noting
that the fraction of tweets on topic 2 (distance learning concerns) increases considerably from 16.95% in the
rst period to 45.94% in the last period. On the other hand, the fraction of tweets on topic 1 (digital support)
decreased during the second period and then grew slightly in the last period. Finally, the number of tweets on
topic 3 (restriction zones) decreased considerably from March 2020 to November 2021.
Limitations
is study has some limitations. Regarding the emotion analysis, a possible limitation is that the number of emo-
tion categories was limited to 828,44, but emotion is a broad concept and may involve up to 27 categories45. Fur-
thermore, misspelled words could not be identied and analyzed in the algorithm. Further limitations concern
the dictionary of sentiments (“lexicon”) developed by Mohammad and Turney28, which maps a list of language
features to emotion intensities:
Only 5 individuals were recruited to annotate a term against each of the 8 primary emotions.
e emotions of a term have been annotated without considering the possible contexts.
(a) k=3(b) k=4
(c) k=7
Figure7. Primer component analysis using average distribution forseveral topic numbers k.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
8
Vol:.(1234567890)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
Although the percentages of agreement were apparently high, interrater reliability statistics were not reported.
Regarding topic analysis, considering unsupervised learning such as DLDA, the primary limitation is some
degree of subjectivity in dening the topic created10. Finally, it is worth noting that the most recent statistics about
social media usage show that approximately 83% of Twitter users worldwide were under age 5046; this implies that
Twitter-based studies generally suer from an underestimation bias in the opinions of people aged 50 and over.
Figure8. Topics and top terms of dierent time slices by DTM. Distributions for some relevant terms in each
topic over time.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
9
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
However, the distance learning topic truly aects the younger population more closely than the older popula-
tion; therefore, the underestimation issue may have a marginal, if any, impact on the results in the present study.
Conclusions and future prospectives
With the aim of studying the opinions and emotions of Italians regarding distance learning, we collected tweets
on this issue and carried out a sentiment analysis using the VADER and syuzhet packages. e results showed
a predominance of negative attitudes. e sentiment analysis shows daily uctuations (Fig.3), mainly due to
continuous updates by the news media and the succession of government decrees to contain the coronavirus.
However, the long-term trend shows an improvement in sentiment until the trend is reversed; attitudes become
positive at the beginning of the 2021–22 school year. Of the highest emotions detected, ‘trust’ was found to be
the main positive emotion, while ‘fear’, ‘sadness’ and ‘anger’ were the top negative emotions. e topic model
identied three topics: (1) requests for support measures for distance learning, (2) concerns about distance learn-
ing and its application, and (3) anxiety about the government decrees introducing red zones and corresponding
restrictions. What emerges clearly is the change over time in the percentage weight of the topics: the concerns
about distance learning assumed an increasing importance to the detriment of the other topics. In the past two
years, the use of distance learning has usurped other learning systems due to the pandemic, inducing sudden,
dramatic and probably irreversible changes in the education process. e use of digital teaching technologies
accelerated and led to a hybrid instructional model that combined remote and face-to-face teaching, named
integrated digital learning. While distance learning has generated and still generates fears and concerns, inte-
grated digital learning has already proven itself more eective than traditional teaching. e positive peak in
time series sentiments started at the beginning of school year 2021–22 (Fig.3) when integrated digital learning
was fully applied in Italy. Further, ongoing technological advancements and the growing experience of students
and teachers could mitigate any concerns related to a return to distance learning following a new pandemic wave
or other crisis. erefore, future studies could investigate how perceptions and opinions about distance learning
will change in the coming years, using sources other than Twitter and combining results of multiple databases.
Received: 16 February 2022; Accepted: 18 May 2022
References
1. Italian Government, DPCM March 11, 2020. Further implementing provisions of the decree-law February 23, 2020, n. 6, contain-
ing urgent measures regarding the containment and management of the epidemiological emergency from COVID-19, applicable
throughout the national territory. https:// www. gazze ttau cia le. it/ eli/ id/ 2020/ 03/ 11/ 20A01 605/ sg (2020). Accessed 4 April 2022.
2. Distance learning solutions. UNESCO. htt ps:// en. un esco. org/ covid 19/ educa tionr espon se/ solut ions (2020). Accessed 4 April 2022.
3. Capone, R. & Lepore, M. From distance learning to integrated digital learning: A fuzzy cognitive analysis focused on engagement,
motivation, and participation during COVID-19 pandemic. Technol. Knowl. Learn.https:// doi. org/ 10. 1007/ s10758- 021- 09571-w
(2021).
4. Kemp, S. Digital 2020: Global digital overview. https:// datar eport al. com/ repor ts/ digit al- 2020- global- digit alove rview (2020).
Accessed 4 April 2022.
5. Zhan, Y., Etter, J.-F., Leischow, S. & Zeng, D. Electronic cigarette usage patterns: A case study combining survey and social media
data. J. Am. Med. Inform. Assoc. 26, 9–18. https:// doi. org/ 10. 1093/ jamia/ ocy140 (2019).
6. Hassanpour, S., Tomita, N., DeLise, T., Crosier, B. & Marsch, L. A. Identifying substance use risk based on deep neural networks
and Instagram social media data. Neuropsychopharmacology 44, 487–494. https:// doi. org/ 10. 1038/ s41386- 018- 0247-x (2019).
7. Rains, S. A., Leroy, G., Warner, E. L. & Harber, P. Psycholinguistic markers of COVID-19 conspiracy tweets and predictors of tweet
dissemination. Health Commun.https:// doi. org/ 10. 1080/ 10410 236. 2021. 19296 91 (2021).
8. He, L. et al. Why do people oppose mask wearing? A comprehensive analysis of U.S. tweets during the COVID-19 pandemic. J.
Am. Med. Inform. Assoc. 28, 1564–1573. https:// doi. org/ 10. 1093/ jamia/ ocab0 47 (2021).
9. Ainley, E., Witwicki, C., Tallett, A. & Graham, C. Using twitter comments to understand people’s experiences of UK health care
during the COVID-19 pandemic: ematic and sentiment analysis. J. Med. Internet Res.https:// doi. org/ 10. 2196/ 31101 (2021).
10. Kwok, S. W. H., Vadde, S. K. & Wang, G. Tweet topics and sentiments relating to COVID-19 vaccination among Australian twitter
users: Machine learning analysis. J. Med. Internet Res. 23, e26953. https:// doi. org/ 10. 2196/ 26953 (2021).
11. Aljabri, M. et al. Sentiment analysis of Arabic tweets regarding distance learning in Saudi Arabia during the COVID-19 pandemic.
Sensors (Basel) 21, 5431. https:// doi. org/ 10. 3390/ s2116 5431 (2021).
Figure9. Dynamic volume of thetopics over time.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
10
Vol:.(1234567890)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
12. Mujahid, M. et al. Sentiment analysis and topic modeling on tweets about online education during COVID-19. Appl. Sci. (Basel)
11, 8438. https:// doi. org/ 10. 3390/ app11 188438 (2021).
13. Asare, A. O., Yap, R., Truong, N. & Sarpong, E. O. e pandemic semesters: Examining public opinion regarding online learning
amidst COVID-19. J. Comput. Assist. Learn. 37, 1591–1605. https:// doi. org/ 10. 1111/ jcal. 12574 (2021).
14. Lee, S. A. & Crunk, E. A. Fear and psychopathology during the COVID-19 crisis: Neuroticism, hypochondriasis, reassurance-
seeking, and coronaphobia as fear factors. Omega (Westport).https:// doi. org/ 10. 1177/ 00302 22820 949350 (2020).
15. Satici, B., Gocet-Tekin, E., Deniz, M. E. & Satici, S. A. Adaptation of the fear of COVID-19 scale: Its association with psychologi-
cal distress and life satisfaction in turkey. Int. J. Ment. Health Addict. 19, 1980–1988. https:// doi. org/ 10. 1007/ s11469- 020- 00294-0
(2021).
16. Duong, C. D. e impact of fear and anxiety of covid-19 on life satisfaction: Psychological distress and sleep disturbance as media-
tors. Pers. Individ. Dier. 178, 110869. https:// doi. org/ 10. 1016/j. paid. 2021. 110869 (2021).
17. L a Rosa, V. L., G ori, A., Faraci, P., Vicario, C. M. & Craparo, G. Traumatic distress, alexithymia, dissociation, and risk of addiction
during the rst wave of COVID-19 in Italy: Results from a cross-sectional online survey on a non-clinical adult sample. Int. J.
Ment. Health Addict.https:// doi. org/ 10. 1007/ s11469- 021- 00569-0 (2021).
18. Biviá-Roig, G. et al. Analysis of the impact of the connement resulting from COVID-19 on the lifestyle and psychological wellbe-
ing of Spanish pregnant women: An internet-based cross-sectional survey. Int. J. Environ. Res. Public Health 17, 5933. https:// doi.
org/ 10. 3390/ ijerp h1716 5933 (2020).
19. Plagg, B., Engl, A., Piccoliori, G. & Eisendle, K. Prolonged social isolation of the elderly during COVID-19: Between benet and
damage. Arch. Gerontol. Geriatr. 89, 104086. https:// doi. org/ 10. 1016/j. archg er. 2020. 104086 (2020).
20. Savci, C., Cil Akinci, A., Yildirim Usenmez, S. & Keles, F. e eects of fear of COVID-19, loneliness, and resilience on the qual-
ity of life in older adults living in a nursing home. Geriatr. Nurs. 42, 1422–1428. https:// doi. org/ 10. 1016/j. gerin urse. 2021. 09. 012
(2021).
21. Commodari, E. & La Rosa, V. L. Adolescents in quarantine during COVID-19 pandemic in Italy: Perceived health risk, beliefs,
psychological experiences and expectations for the future. Front. Psychol. 11, 559951. https:// doi. org/ 10. 3389/ fpsyg. 2020. 559951
(2020).
22. Rania, N. & Coppola, I. e fear of contagion and the attitude toward the restrictive measures imposed to face COVID-19 in Italy:
e psychological consequences caused by the pandemic one year aer it began. Front. Psychol. 13, 805706. https:// doi. org/ 10.
3389/ fpsyg. 2022. 805706 (2022).
23. Tumasjan, A., Sprenger, T., Sandner, P. & Welpe, I. Predicting elections with twitter: What 140 characters reveal about political
sentiment. In Proc. Fourth Int. AAAI Conf. Weblogs Soc. Media Predict., vol.10 (2010).
24. Oyebode, O., Orji, R. Social. & media and sentiment analysis: e Nigeria presidential election,. In 2019 IEEE 10th Annual Infor-
mation Technology. Electronics and Mobile Communication Conference (IEMCON) 2019. https:// doi. org/ 10. 1109/ IEMCON. 2019.
89361 39 (IEEE, 2019).
25. Budiharto, W. & Meiliana, M. Prediction and analysis of Indonesia presidential election from twitter using sentiment analysis. J.
Big Data 5, 2. https:// doi. org/ 10. 1186/ s40537- 018- 0164-1 (2018).
26. Bird, S., Klein, E. & Loper, E. Natural Language Processing with Python (O’Reilly Media, ***, 2009).
27. Hutto, C. & Gilbert, E. Vader: A parsimonious rule-based model for sentiment analysis of social media text. In Proceedings of the
8th International Conference on Weblogs and Social Media, ICWSM 2014 (2015).
28. Mohammad, S. & Turney, P. Emotions evoked by common words and phrases: Using mechanical turk to create an emotion lexicon.
In Proceedings of the NAACL HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text
(LA, California, 2010).
29. Jockers, M. L. Syuzhet: Extract sentiment and plot arcs from text. GitHub. https:// github. com/ mjock ers/ syuzh et (2015). Accessed
10 January 2022.
30. Blei, D. M., Ng, A. Y., Jordan, M. I. & Laerty, J. Latent Dirichlet allocation. J. Mach. Learn. Res. 3, 993–1022 (2003).
31. Lee, J. et al. Ensemble modeling for sustainable technology transfer. Sustainability 10, 22–78. https:// doi. org/ 10. 3390/ su100 72278
(2018).
32. Blei, D.M. & Laerty, J.D. Dynamic topic models. In Proceedings of the 23rd International Conference on Machine Learning—ICML
’06. https:// doi. org/ 10. 1145/ 11438 44. 11438 59 (ACM Press, 2006).
33. Řehůrek, R. & Sojka, P. In Proceedings of LREC 2010 Workshop New Challenges for NLP Frameworks (Valletta, Malta).
34. Sievert, C. & Shirley, K. LDAvis: A method for visualizing and interpreting topics. In Proceedings of the Workshop on Interactive
Language Learning, Visualization, and Interfaces, 63–70, (Association for Computational Linguistics, 2014). https:// doi. org/ 10.
3115/ v1/ W14- 3110. https:// aclan tholo gy. org/ W14- 3110.
35. Montemurro, N. e emotional impact of COVID-19: From medical sta to common people. Brain Behav. Immun. 87, 23–24.
https:// doi. org/ 10. 1016/j. bbi. 2020. 03. 032 (2020).
36. Saricali, M., Satici, S. A., Satici, B., Gocet-Tekin, E. & Griths, M. D. Fear of COVID-19, mindfulness, humor, and hopelessness:
A multiple mediation analysis. Int. J. Ment. Health Addict.https:// doi. org/ 10. 1007/ s11469- 020- 00419-5 (2020).
37. Satici, S. A., Kayis, A. R., Satici, B., Griths, M. D. & Can, G. Resilience, hope, and subjective happiness among the Turkish popula-
tion: Fear of COVID-19 as a mediator. Int. J. Ment. Health Addict.https:// doi. org/ 10. 1007/ s11469- 020- 00443-5 (2020).
38. Deniz, M. E. Self-compassion, intolerance of uncertainty, fear of COVID-19, and well-being: A serial mediation investigation.
Pers. Individ. Dier. 177, 110824. https:// doi. org/ 10. 1016/j. paid. 2021. 110824 (2021).
39. Daly, M. & Robinson, E. Psychological distress and adaptation to the COVID-19 crisis in the united states. J. Psychiatr. Res. 136,
603–609. https:// doi. org/ 10. 1016/j. jpsyc hires. 2020. 10. 035 (2021).
40. Lee, C. M., Cadigan, J. M. & R hew, I. C. Increases in loneliness among young adults during the COVID-19 pandemic and associa-
tion with increases in mental health problems. J. Adolesc. Health 67, 714–717. https:// doi. org/ 10. 1016/j. jadoh ealth. 2020. 08. 009
(2020).
41. Ye, B. et al. Stressors of COVID-19 and stress consequences: e mediating role of rumination and the moderating role of psy-
chological support. Child. Youth Serv. Rev. 118, 105466. https:// doi. org/ 10. 1016/j. child youth. 2020. 105466 (2020).
42. Mouchet-Mages, S. & Baylé, F. J. Sadness as an integral part of depression. Dialogues Clin. Neurosci. 10, 321–327. https:// doi. org/
10. 31887/ DCNS. 2008. 10.3/ smmag es (2008).
43. Abadi, D., Arnaldo, I. & Fischer, A. Anxious and angry: Emotional responses to the COVID-19 threat. Front. Psychol. 12, 676116.
https:// doi. org/ 10. 3389/ fpsyg. 2021. 676116 (2021).
44. Plutchik, R. A general psychoevolutionary theory of emotion. In eories of Emotion 3–33. https:// doi. org/ 10. 1016/ b978-0- 12-
558701- 3. 50007-7 (Elsevier, 1980).
45. Cowen, A. S. & Keltner, D. Self-report captures 27 distinct categories of emotion bridged by continuous gradients. Proc. Natl. Acad.
Sci. U. S. A. 114, E7900–E7909. https:// doi. org/ 10. 1073/ pnas. 17022 47114 (2017).
46. Statista. Distribution of twitter users worldwide as of April 2021, by age group. https:// www. stati sta. com/ stati stics/ 283119/ age- distr
ibuti on- of- global- twitt er- users/ (2021). Accessed 7 May 2022.
Author contributions
e two authors contributed equally to this work.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
11
Vol.:(0123456789)
Scientic Reports | (2022) 12:9163 | https://doi.org/10.1038/s41598-022-12915-w
www.nature.com/scientificreports/
Competing interests
On behalf of all authors, the corresponding author states that there is no conict of interest. e datasets used
and analyzed during this study are historical Twitter data purchased through the https:// www. track myhas htag.
com/ service which provides data in line with Twitter’s T & Cs. ese datasets are available from the correspond-
ing author upon reasonable request.
Additional information
Correspondence and requests for materials should be addressed to L.S.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access is article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. e images or other third party material in this
article are included in the articles Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
© e Author(s) 2022
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The pandemic nature of COVID-19 has caused major changes in health, economy, and society globally. Albeit to a lesser extent, contingent access to shops and places to socialize the imposition of social distancing and the use of indoor masks is measures still in force today (more than a year after the start of the pandemic), with repercussions on economic, social, and psychological levels. The fear of contagion, in fact, has led us to be increasingly suspicious and to isolate ourselves from the remainder of the community. This has had repercussions on the perception of loneliness, with significant psychological consequences, such as the development of stress, anxiety, and, in extreme cases, depressive symptoms. Starting from these assumptions, this research was developed with the aim of deepening the perceptions that the participants have of their own mental health, loneliness, fear linked to contagion, and attitudes toward imposed social distancing. In particular, we wanted to analyze whether there is a relationship between perceived fear and the perceived level of mental health, loneliness, and attitude toward social distancing. Finally, we wanted to analyze whether there are differences related to gender, age, marital status, current working mode, and educational qualifications. The research, performed after the diffusion of the vaccination in Italy, lasted 14 days. The participants were 500 Italians who voluntarily joined the study and were recruited with random cascade sampling. The research followed a quantitative approach. The analyzed data, from participants residing throughout the national territory, allow us to return the picture of the perceptions that Italians have of the fear of contagion, of their level of mental health, of loneliness and of their attitude toward social distancing. In particular, the data show that fear of COVID-19 is an emotional state experienced by the entire population and that young people have suffered more from loneliness and have been less inclined to accept the imposed social distancing. The data that emerged should make policymakers reflect on the need to find functional strategies to combat COVID-19 or other health emergency crises whose effects do not affect the psychological wellbeing of the population.
Article
Full-text available
This work focuses on Distance Learning during the COVID-19 pandemic to improve students’ motivation, participation, and engagement, trying to contain the drop-out phenomenon. Distance Learning at the time of COVID-19 is an educational methodology and it can be considered the only occasion to keep an educational connection between students and teachers. Experimentation in a Mathematics STEM class was carried out evaluating the impact of Distance Learning on students’ levels of motivation, participation, and engagement, computed through a Fuzzy Cognitive Map. Specifically, it was performed on some affective and interaction parameters derived from using an adaptive e-learning platform and from the answers of a semi-structured questionnaire. The results, which have been analysed through Technological Pedagogical Content Knowledge and Instrumental Genesis theories, show on one hand that Distance Learning is valid as an additional and support methodology but, on the other hand, they highlight the ineffectiveness of completely remote teaching. Therefore, a teaching method that integrates moments of distance teaching with activities carried out in the presence, in the classroom, or in other university environments, is hoped to be used as soon as the emergency is over: a mix of styles, a fluid flow of knowledge between the physical classroom and the virtual classroom. We will call this Integrated Digital Learning.
Article
Full-text available
This study aimed to evaluate the fear of COVID-19, loneliness, resilience, and quality of life levels in older adults in a nursing home during the pandemic, and the effects of these variables and descriptive characteristics on their quality of life. Data were collected using a participant information form, the Mini Mental State Exam (MMSE), the Brief Resilience Scale, the Fear of COVID-19 Scale (FCV-19S), the Loneliness Scale for Elderly (LSE), and the World Health Organization Quality of Life-BREF Turkish Version (WHOQOL-BREF-TR). Regarding the WHOQOL-BREF-TR scale, being male, history of chronic disease, MMSE score, and the changes in sleep pattern significantly affected the physical dimension (R²=0.353, p<0.01) while age, and MMSE and LSE scores significantly affected the psychological dimension (R²=0.364, p<0.01). Also, the MMSE, FCV-19S, and LSE scores significantly affected the social relations dimension (R²=0.234, p<0.01) while MMSE, FCV-19S, and LSE scores significantly affected the environmental dimension (R²=0.351, p<0.01).
Article
Full-text available
Amid the worldwide COVID-19 pandemic lockdowns, the closure of educational institutes leads to an unprecedented rise in online learning. For limiting the impact of COVID-19 and obstructing its widespread, educational institutions closed their campuses immediately and academic activities are moved to e-learning platforms. The effectiveness of e-learning is a critical concern for both students and parents, specifically in terms of its suitability to students and teachers and its technical feasibility with respect to different social scenarios. Such concerns must be reviewed from several aspects before e-learning can be adopted at such a larger scale. This study endeavors to investigate the effectiveness of e-learning by analyzing the sentiments of people about e-learning. Due to the rise of social media as an important mode of communication recently, people’s views can be found on platforms such as Twitter, Instagram, Facebook, etc. This study uses a Twitter dataset containing 17,155 tweets about e-learning. Machine learning and deep learning approaches have shown their suitability, capability, and potential for image processing, object detection, and natural language processing tasks and text analysis is no exception. Machine learning approaches have been largely used both for annotation and text and sentiment analysis. Keeping in view the adequacy and efficacy of machine learning models, this study adopts TextBlob, VADER (Valence Aware Dictionary for Sentiment Reasoning), and SentiWordNet to analyze the polarity and subjectivity score of tweets’ text. Furthermore, bearing in mind the fact that machine learning models display high classification accuracy, various machine learning models have been used for sentiment classification. Two feature extraction techniques, TF-IDF (Term Frequency-Inverse Document Frequency) and BoW (Bag of Words) have been used to effectively build and evaluate the models. All the models have been evaluated in terms of various important performance metrics such as accuracy, precision, recall, and F1 score. The results reveal that the random forest and support vector machine classifier achieve the highest accuracy of 0.95 when used with Bow features. Performance comparison is carried out for results of TextBlob, VADER, and SentiWordNet, as well as classification results of machine learning models and deep learning models such as CNN (Convolutional Neural Network), LSTM (Long Short Term Memory), CNN-LSTM, and Bi-LSTM (Bidirectional-LSTM). Additionally, topic modeling is performed to find the problems associated with e-learning which indicates that uncertainty of campus opening date, children’s disabilities to grasp online education, and lagging efficient networks for online education are the top three problems.
Article
Full-text available
The current coronavirus disease 2019 (COVID-19) pandemic elicits a vast amount of anxiety. In the current study, we investigated how anxiety related to COVID-19 is associated with support for and compliance with governmental hygiene measures, and how these are influenced by populist attitudes, anger at the government, and conspiracy mentalities. We conducted an online survey in April 2020 in four different countries (Germany, the Netherlands, Spain, and the UK; N = 2,031) using a cross-sectional design. Results showed that (1) anxiety related to COVID-19 is associated with conspiracy beliefs, anger at the government, and populist attitudes, and (2) support for and compliance with hygiene measures are both positively predicted by anxiety related to COVID-19; however, (3) support for hygiene measures is also predicted by populist attitudes and negatively by conspiracy mentalities, whereas compliance with hygiene measures is more strongly predicted by anger at transgressors (anger at people transgressing the hygiene measures). Consequently, although anxiety related to COVID-19 concerns the health of individual people, it also has political and social implications: anxiety is associated with an increase in anger, either at transgressors or the government.
Article
Full-text available
The COVID-19 pandemic has greatly impacted the normal life of people worldwide. One of the most noticeable impacts is the enforcement of social distancing to reduce the spread of the virus. The Ministry of Education in Saudi Arabia implemented social distancing measures by enforcing distance learning at all educational stages. This measure brought about new experiences and challenges to students, parents, and teachers. This research measures the acceptance rate of this way of learning by analysing people's tweets regarding distance learning in Saudi Arabia. All the tweets analysed were written in Arabic and collected within the boundary of Saudi Arabia. They date back to the day that the distance learning announcement was made. The tweets were pre-processed, and labelled positive, or negative. Machine learning classifiers with different features and extraction techniques were then built to analyse the sentiment. The accuracy results for the different models were then compared. The best accuracy achieved (0.899) resulted from the Logistic regression classifier with unigram and Term Frequency-Inverse Document Frequency as a feature extraction approach. This model was then applied on a new unlabelled dataset and classified to different educational stages; results demonstrated generally positive opinions regarding distance learning for general education stages (kindergarten, intermediate, and high schools), and negative opinions for the university stage. Further analysis was applied to identify the main topics related to the positive and negative sentiment. This result can be used by the Ministry of Education to further improve the distance learning educational system.
Article
Full-text available
The current educational disruption caused by the COVID-19 pandemic has fuelled a plethora of investments and the use of educational technologies for Emergency Remote Learning (ERL). Despite the significance of online learning for ERL across most educational institutions, there are wide mixed perceptions about online learning during this pandemic. This study, therefore, aims at examining public perception about online learning for ERL during COVID-19. The study sample included 31,009 English language Tweets extracted and cleaned using Twitter API, Python libraries and NVivo, from 10 March 2020 to 25 July 2020, using keywords: COVID-19, Corona, e-learning, online learning, distance learning. Collected tweets were analysed using word frequencies of unigrams and bigrams, sentiment analysis, topic modelling, and sentiment labeling, cluster, and trend analysis. The results identified more positive and negative sentiments within the dataset and identified topics. Further, the identified topics which are learning support, COVID-19, online learning, schools, distance learning, e-learning, students, and education were clustered among each other. The number of daily COVID-19 related cases had a weak linear relationship with the number of online learning tweets due to the low number of tweets during the vacation period from April to June 2020. The number of tweets increased during the early weeks of July 2020 as a result of the increasing number of mixed reactions to the reopening of schools. The study findings and recommendations underscore the need for educational systems, government agencies, and other stakeholders to practically implement online learning measures and strategies for ERL in the quest of reopening of schools.
Article
Full-text available
This study aimed to explore the prevalence of post-traumatic distress, alexithymia, dissociation, and addictive behaviors during the stressful situation of the COVID-19 pandemic. It also aimed to determine whether trauma, alexithymia, and dissociation can effectively predict the risk of addiction in non-clinical subjects during the COVID-19 pandemic. Two hundred nineteen subjects completed a web survey during the first COVID-19 lockdown in Italy, including the Impact of Event Scale-Revised (IES-R), the Dissociative Experience Scale-II (DES-II), and the Addictive Behavior Questionnaire (ABQ). Females reported higher levels of COVID-19-related traumatic stress than males (p = 0.009). A greater fear of getting COVID-19 was associated with significantly high IES-R scores (p < 0.0005). IES-R total score was significantly lower in the “not internet-addicted” group than that in the “internet-addicted” group (p < 0.0005). Furthermore, DES-II total score was significantly higher in the “internet-addicted” group than that in the “non internet-addicted” group (p < 0.0005). No statistically significant score differences were highlighted in the “alcohol” group. Future research with longitudinal studies and larger samples will have to clarify whether trauma, alexithymia, and dissociation can effectively predict the risk of addiction in non-clinical subjects during the COVID-19 pandemic.
Article
Background: The Covid-19 pandemic has led to changes in health service utilization patterns and a rapid rise in care being delivered remotely. There has been little published research examining patients' experiences of accessing remote consultations since Covid-19. Such research is important as remote methods for delivering some care may be maintained in the future. Objective: To use content from Twitter to understand discourse around health and care delivery in the UK as a result of Covid-19, in particular Twitter users' views on and attitudes to care being delivered remotely. Methods: Tweets posted from the UK between January 2018 and October 2020 were extracted using the Twitter API. 1,408 tweets across three search terms were extracted into Excel. 161 tweets were removed following de-duplication, and 610 were identified as irrelevant to the research question. Relevant tweets (n=637) were coded into categories using NVivo software, and assigned a positive, neutral, or negative sentiment. To examine views of remote care over time, the coded data was imported back into Excel so that each tweet was associated with both a theme(s) and sentiment. Results: The volume of tweets on remote care delivery increased markedly following the Covid-19 outbreak. Five main themes were identified in the tweets: access to remote care (n=267), quality of remote care (n=130), anticipation of remote care (n=39), online booking and asynchronous communication (n=85) and publicising changes to services or care delivery (n=160). Mixed public attitudes and experiences to the changes in service delivery were found. The proportion of positive tweets regarding access to, and quality of, remote care was higher in the immediate period following the Covid-19 outbreak (March-May 2020) when compared to the time before the Covid-19 onset, and the time when restrictions from the first lockdown eased (June-October 2020). Conclusions: Using Twitter data to address our research questions proved beneficial for providing rapid access to Twitter users' attitudes to remote care delivery at a time when it would have been difficult to conduct primary research due to Covid-19. It allowed us to examine the discourse on remote care over a relatively long period and explore shifting attitudes of Twitter users at a time of rapid changes in care delivery. The mixed attitudes towards remote care highlights the importance that patients have a choice over the type of consultation that best suits their needs, and that the increased use of technology for delivering care does not become a barrier for some. The finding that overall sentiment about remote care was more positive in the early stages of the pandemic but since declined emphasises the need for a continued examination of people's preference, particularly if remote appointments are likely to remain central to healthcare delivery. Clinicaltrial:
Article
The adoption of conspiracy theories about COVID-19 has been fairly widespread among the general public and associated with the rejection of self-protective behaviors. Despite their significance, however, a gap remains in our understanding of the underlying characteristics of messages used to disseminate COVID-19 conspiracies. We used the construct of resonance as a framework to examine a sample of more than 1.8 million posts to Twitter about COVID-19 made between April and June 2020. Our analyses focused on the psycholinguistic properties that distinguish conspiracy theory tweets from other COVID-19 topics and predict their spread. COVID-19 conspiracy tweets were distinct and most likely to resonate when they provided explanations and expressed negative emotions. The results highlight the sensemaking functions served by conspiracy tweets in response to the profound upheaval caused by the pandemic.