engleski

Sentiment Analysis of the Infographics on Social Media

Today, the Internet is the main source of information for both users and businesses. Communication and exchange of information are two- way, but results are often obtained unstructured, multimodal, informal, high- dimensional, and noisy. This is especially pronounced with data collected from social media which are an indispensable marketing tool. A large number of active users allows you to quickly spread the sent message, while also providing almost instant feedback. However, not all forms of sent messages are noticed and spread equally. Therefore, in this paper, a sentimental analysis of the users’ reactions was made. Multimodal data were used to collect users’ reactions: text publication, image publication, and infographics on the topic of ecology. All three types of publications are posted on three social media, Facebook, Twitter, and LinkedIn. Naïve Bayes, Decision Tree, Random Forest, Support Vector Machine, and Neural Network Algorithms were used to automatically assess sentiment as positive, negative, or neutral. The results thus given show the best overall performance for Decision Tree classifier on all types of users’ reactions. Then follows Random Forest and finally Naïve Bayes classifier. Also, the obtained results show data linearity from which it can be concluded that on whatever social media it is published, infographics certainly has a leading position when it comes to spotting and getting users’ reactions, especially positive ones.

sentiment analysis ; infographics ; social media

nije evidentirano