Amanatidou, Aglaia

2022-06-16T10:37:51Z

https://repository.ihu.edu.gr//xmlui/handle/11544/29983

Sentiment analysis

Sentiment analysis of cosmetic industry YouTube video campaigns

2022-05-12

free

School of Science and Technology, MSc in e-Business and Digital Marketing

This dissertation was written as a part of the MSc in e-Business and Digital Marketing at the International Hellenic University. The main goal of the study is to analyze YouTube video comments from cosmetic industry channels, in order to classify them into positive, negative or neutral. Marketing and communication experts spend millions of dollars every year to create YouTube campaigns that make an impact. In this study, the general sentiment in beauty industry, different brands and content types are analyzed to understand if campaigns have positive, negative or neutral impact to users. This analysis could help digital marketing experts of beauty industry to create more useful and successful campaigns. In this study, 187 videos and 189,981 comments have been analyzed by TextBlob, which is a natural language processing machine learning library. Our findings show that the general sentiment in beauty industry YouTube campaigns is positive, across all brands and content types. The results, additionally, show that the model had 59.21% accuracy, 56.95% precision and 78.89% recall compared to the sentiment score provided by a human.

Tzafilkou, Aikaterini

ihu