Sentiment Analysis of Folk Art Social Campaigns

Folk art is a deep method of social communication, which shows cultural identity, emotions of the community, and collective stories. The use of folk art in social campaigns is growing in the digital age as the means of triggering the feelings of empathy and reinforcing the awareness of the culture, although there is a lack of systemic assessment of the emotional response. The paper will investigate the combination of sentiment analysis strategies to assess the reactions of the public to folk art related social campaigns. An extensive database was assembled on social media and campaign archives of folk-art inspired textual and visual materials. To clean up textual data, preprocessing was done by means of tokenization, removal of stop words and normalization. Lexicon-based and machine learning models (SVM, Random Forest) and deep learning models (CNN, LSTM, and BERT) were used to classify sentiments. The higher order methods of feature extraction (TF-IDF, Word2vec and embedding BERT) were implemented with a view to augmenting semantic knowledge. The results of the analysis showed that there are high correlations between the cultural symbolism and emotional involvement, showing that folk motifs and regional idioms provoke more positive emotions than generic campaign designs. Results highlight the point that in addition to enhancing message resonance folk art can also be used to fill socio-cultural gaps by enhancing communication that is emotionally based

Deep learning · Empathy · Feeling · Imitation · Sentiment analysis · Social media · Word2vec · Aesthetic Perception and Analysis · Cultural and Communication Design Research · Public Spaces through Art

• Fake online reviews

  Open Access•Yuanyuan Wu, Eric W T Ngai et al.•Decision Support Systems•2020

  Add to list