Emotion lexicons are useful in research across various disciplines, but the availability of such resources remains limited for most languages. While existing emotion lexicons typically comprise words, it is a particular meaning of a word (rather than the word itself) that conveys emotion. To mitigate this issue, we present the Emotion Meanings dataset, a novel dataset of 6000 Polish word meanings. The word meanings are derived from the Polish wordnet (plWordNet), a large semantic network interlinking words by means of lexical and conceptual relations. The word meanings were manually rated for valence and arousal, along with a variety of basic emotion categories (anger, disgust, fear, sadness, anticipation, happiness, surprise, and trust). The annotations were found to be highly reliable, as demonstrated by the similarity between data collected in two independent samples: unsupervised ( n = 21,317) and supervised ( n = 561). Although we found the annotations to be relatively stable for female, male, younger, and older participants, we share both summary data and individual data to enable emotion research on different demographically specific subgroups. The word meanings are further accompanied by the relevant metadata, derived from open-source linguistic resources. Direct mapping to Princeton WordNet makes the dataset suitable for research on multiple languages. Altogether, this dataset provides a versatile resource that can be employed for emotion research in psychology, cognitive science, psycholinguistics, computational linguistics, and natural language processing.
Introduction: Vaccinations are referred to as one of the greatest achievements of modern medicine. However, their effectiveness is also constantly denied by certain groups in society. This results in an ongoing dispute that has been gradually moving online in the last few years due to the development of technology. Our study aimed to utilize social media to identify and analyze vaccine-deniers’ arguments against child vaccinations.
Method: All public comments posted to a leading Polish vaccination opponents’ Facebook page posted between 01/05/2019 and 31/07/2019 were collected and analyzed quantitatively in terms of their content according to the modified method developed by Kata (Kata, 2010). Sentiment analysis was also performed.
Results: Out of 18,685 comments analyzed, 4,042 contained content covered by the adopted criteria: conspiracy theories (28.2%), misinformation and unreliable premises (19.9%), content related to the safety and effectiveness of vaccinations (14.0%), noncompliance with civil rights (13.2%), own experience (10.9%), morality, religion, and belief (8.5%), and alternative medicine (5.4%). There were also 1,223 pro-vaccine comments, of which 15.2% were offensive, mocking, or non-substantive. Sentiment analysis showed that comments without any arguments as well as those containing statements about alternative medicine or misinformation were more positive and less angry than comments in other topic categories.
Conclusions: The large amount of content in the conspiracy theory and misinformation categories may indicate that authors of such comments may be characterized by a lack of trust in the scientific achievements of medicine. These findings should be adequately addressed in vaccination campaigns.
@article{doi:10.1080/21645515.2020.1850072,
author = {Krzysztof Klimiuk and Agnieszka Czoska and Karolina Biernacka and Łukasz Balwicki},
title = {Vaccine misinformation on social media – topic-based content and sentiment analysis of Polish vaccine-deniers’ comments on Facebook},
journal = {Human Vaccines \& Immunotherapeutics},
volume = {0},
number = {0},
pages = {1-10},
year = {2021},
publisher = {Taylor & Francis},
doi = {10.1080/21645515.2020.1850072},
note ={PMID: 33517844},
The topic of vaccines, and specifically the differences between them, arouses great emotion. Analysis of Google search phrases shows that the public is still looking for information on the differences between preparations of different manufacturers. “Which vaccine to choose?”, “Moderna or Pfizer?” – are just some of them. This time we took to the workshop conversations of Internet users about five preparations. We explored the emotions and sentiment between them to see how they differ.
Pfizer or Moderna, or which vaccine to choose?
The material analyzed by our team represents a total of nearly 190,000 mentions of five vaccine manufacturers between January 1 and May 6, 2021. Content about specific manufacturers was filtered out of discussions about COVID-19 vaccines overall. Using the SentiTool tool, it only took a few minutes to get the full results.
Preparations were most often discussed by Internet users under articles on news sites. The second place in terms of discussion sites was Twitter, followed by Facebook. On news portals, Internet users discussed AstraZeneca most often. This is also where a lot of opinions about the Sputnik-V vaccine from Russia were reported online. On Twitter and Facebook, but also in other places, it was Pfizer. These two vaccines are among the most commented on.
The graph of discussion dynamics (monthly data) shows that the AstraZeneca controversy has had its effect. This vaccine was discussed very intensively – most often in March this year. Johnson&Johnson was the least discussed. It is worth noting that in a significant number of opinions there were more names of preparations.
COVID-19 vaccines – a comparison. Sentiment and emotion
When looking at the differences in sentiment of opinion of a given vaccine versus the others, two extremes are clearly visible. Johnson’s was the most positively commented on during the study period. In contrast, the most negative statements were reported around Sputnik-V. The controversy around vaccine reactions after AstraZeneca’s vaccine ultimately did not impair the online image. More unfavorable content was seen from Pfizer, which may come as a surprise.
DIFFERENCES IN OPINION SENTIMENT BETWEEN VACCINES DURING THE STUDY PERIOD
Differences in the intensity of negative and positive sentiment inopinions on COVID-19 formulations. The graph shows how sentiment differed for each vaccine from the other four
This preparation is one of the most popular among Poles and it is with this preparation that citizens most often want to be vaccinated. This may have been influenced by the co-occurrence of other preparation names together with Pfizer. This is because we are primarily examining the general climate of online discussion around the products of pharmaceutical companies. Thus the graph below shows how opinions on one preparation could have influenced the discussion climate around another preparation on the basis of co-occurrence. Moderna and AstraZeneca co-occurred most frequently with Pfizer.
Co-occurrence analysis of formulations in online reviews
The chart below breaks the positive and negative sentiment into eight base emotions for an even closer look at vaccine opinions. Sputnik-V differed from the other formulations with a high dose of disgust (about 13% more than the others) from internet users, while Pfizer differed with fear and sadness. AstraZeneca reported a slight increase in confidence and expectation. With the same emotions, single-dose Johnson differed from the rest most on the plus side (with strong decreases in negative emotions relative to competitors). As for Moderna, there is little difference from the other four formulations against COVID-19.
The above analysis is illustrative and informative. It does not reflect the full results of the study.
In this article we extend a WordNet structure with relations linking synsets to Desikan’s brain regions. Based on lexicographer files and WordNet Domains the mapping goes from synset semantic categories to behavioural and cognitive functions and then directly to brain lobes. A human brain connectome (HBC) adjacency matrix was utilised to capture transition probabilities between brain regions. We evaluated the new structure in several tasks related to semantic similarity and emotion processing using brain-expanded Princeton WordNet (207k LUs) and Polish WordNet (285k LUs, 30k annotated with valence, arousal and 8 basic emotions). A novel HBC vector representation turned out to be significantly better than proposed baselines. URL: https://www.sciencedirect.com/science/article/pii/S0306457321000388
Emotion analysis is the primary (next to sentiment and emotional arousal analysis, of course) function of the science-based Sentimenti tools we have been developing for many years. With the new year our solution has grown in value and can handle perfectly not only Polish but also seventeen other languages. The following study proves that we do it with almost equal efficiency.
Many languages, one tool. Sentimenti already handles 18 languages
In a summary posted on our blog at the beginning of 2021, specifically on January 4, we reported on new challenges. The Sentimenti team has not settled on its laurels and is working hard to develop and improve the tools. The result? We can now analyze large collections of texts in… 18 languages! However, in order to prove (not only to ourselves) that we are doing this effectively, we decided to analyze one solid text translated into many languages, and then check the emotions, sentiment and emotional arousal it contains.
To raise the bar for our tools, we took a rather funny and often sarcasm-laden text. These requirements were met perfectly by the content of Maysoon Zayid’s speech during TEDWomen in December 2013. Zayid is a writer, actress, comedian and co-founder of New York’s Arab-American Comedy Festival. Read/watch the entire piece below, or: HERE.
Thanks to the analysis with Sentimenti tools of the content of the speech in Polish, English, German and Spanish it has been confirmed that the results of the intensity of emotions are very similar. This proves high effectiveness. But that is not all! Our systems perfectly detected sarcasm and irony contained in the speech, which at times was indeed not lacking.
Emotion analysis versus sarcasm, humor, and irony. How effective?
“I have infantile cerebral palsy. I shake all the time.” – informs the recipient Maysoon Zayid at the very beginning of her unusual, comical speech. The content of her speech is not only sarcasm, humor and irony. Above all, the writer does it with great grace. As an aside, it is worth mentioning that her speech is mainly dedicated to the fact that each of us can do anything if we want to.
She herself has walked the red carpet, starred in a movie with Adam Sandler and worked with many other artists. She has also traveled the world, and the number of things she does every day would put many a healthy person to shame.
Returning to the analysis, however, the content study thus confirmed all the theses posed in the introduction. The percentage of positive sentiment, as well as the sum of the percentages of positive versus negative emotions, are higher. The text emotionally aroused mainly with positive emotions. The emotive data (on the scale), in turn, indicated the actual emotions evoked in the viewer.
This happened despite the strongly humorous style of speech throughout the speech. The emotion analysis and its results in the aforementioned languages are very similar to each other. The subtle differences (which we show in the graph below) between them are more due to the varying verbal volume of the same (in semantic terms) sentences. These differences did not exceed the value of 3%.
One content, multiple languages. For what purpose?
Analysis of emotion, sentiment, and emotional arousal in content has wide applications. Studying the same qualitative parameters in content in multiple languages offers even more possibilities. It is not only the analysis of web publications, extensive excerpts from books or magazines or chats with customers. Thanks to multilingual analysis we have an opportunity to get to know the attitude of the Internet towards various issues without language barrier.
The Internet is a huge space where people share opinions about everything. Internet users comment on practically every topic no matter where they are from. Therefore, we cannot remain indifferent to the potential that lies dormant here. Development of tools to analyze new languages is simply a market requirement.
Monitoring the Internet is a necessity and daily work for many companies, and researching emotions, sentiment and excitement on many markets (not only Polish) gives even more information without which it is difficult to do business.
Sentimenti tools have many applications in various fields – from e-commerce to politics, from image crisis research to customer experience. It’s also an excellent, already multilingual, research tool for various types of studies.
Additional perks? Sentimenti’s tools “read” data as if it were being done by the 20,000+ people who participated in Sentimenti’s research and creation. They also do it much faster than a manual survey. Explore our OFFER.
Używamy plików cookie, aby zapewnić prawidłowe działanie naszej strony internetowej i usług. Te pliki cookie są niezbędne i są ustawiane automatycznie. Chcielibyśmy również używać analitycznych plików cookie, aby zrozumieć, w jaki sposób wchodzisz w interakcję z naszą witryną i marketingowymi plikami cookie, aby zapewnić spersonalizowane treści i reklamy.
Klikając „Akceptuję”, wyrażasz zgodę na używanie przez nas plików cookie zarówno do podstawowych, jak i nieistotnych celów. Kliknij „Ustawienia”, aby wybrać typy plików cookie, które chcesz włączyć lub wyłączyć. Więcej informacji znajdziesz w naszej Polityce prywatności RODO.
Przeczytaj naszą politykę cookie
