Visual sentiment analysis in social media

Asst. Prof. Dr. Heba Jabbar Aleqabie
Dunya Jasim Mohammed

Faculty of Computer Science and Information Technology
Visual material is an efficient way to transmit information, including clues about thoughts and emotions, as represented by the adage, “A picture is worth a thousand words.” In diverse application areas, such as instruction, entertainment, advertising, and journalism, such clues indicating the emotions and attitudes of the photographers may elicit comparable sensations from the viewer and aid in comprehending visual material beyond semantic notions.
Due to the social media platforms being so prominent in today’s culture and network connections to the Internet that may be accessed at any time, more than 2 billion individuals, or around one-third of the world’s entire population, utilize these platforms at any given moment. People increasingly post information, engage in events, and express their thoughts in a number of sectors, from products and services to public opinion, as a result of the growth and prosperity of mobile devices, as well as the emergence of the web and flexibility and convenience, so visual analysis will be critical in determining the mood of viewers and members of these platforms.
Visual analysis may define as a method of appreciating art that emphasizes the visual features of a work, such as the relationships between its colour, line, texture, and dimensions. The strictest definition of the word characterizes it as “a description and justification of the visual structure.” Visual analysis, however, may also be used to determine the decisions the artist made when producing the image and to get a better comprehension of how the formal components of an artwork express concepts, content, or meaning.
The significance of visual social media content
Across practically all internet platforms, videos and photographs have shown to be more interesting than just text-based submissions. LinkedIn posts with photos receive 98% more comments on average, while tweets with visual information are three times more likely to receive interaction.
Carousel posts have the greatest engagement rate on Instagram. According to Statista statistics from 2022, the average engagement rate for any Instagram business page post is 1.94%. Carousel posts, on the other hand, have a 3.15% engagement rate.
These types of statistics influence marketers to claim that videos are the most effective form of social media marketing. Videos were the most often mentioned sort of content in a 2021 Statista study in which 54% of marketers claimed that they are useful for achieving social media marketing objectives. resulting in the significance of visual material.
What is Visual sentiment analysis used for?
It is used for Emotional analysis that has been put to use in a wide variety of applications, including. Studying a person’s psychology by looking at their social media profile image posts, and examining visual content to spot inappropriate material for kids, such as violence or anything that encourages bad behaviour, as well as sexual material. Movie posters are examined and described in order to deduce their meaning before the film is released. Examine ideas, replies, and interactions about societal concerns utilizing the visual material offered by social networking sites, helping with picture recovery and providing personalized recommendations. Applications related to the entertainment industry.
In conclusion, visual material is an effective means of conveying information, including clues about ideas and feelings. Social media platforms are widely used, with over 2 billion people utilizing them at any given time. Visual analysis is a way of assessing art that highlights a work’s visual characteristics, such as colour, line, texture, and proportions. It may also be used to identify the artist’s decisions in making the picture and to comprehend how the formal features of an artwork transmit ideas, substance, or meaning. On social media, visual material is more engaging than text-based posts, with LinkedIn posts with photos receiving 98% more comments and tweets with visual content receiving the greatest engagement rate. Emotional data is analyzed via visual sentiment analysis.
References
S. Z. Hassan, K. Ahmad, A. Al-Fuqaha, and N. Conci, “Sentiment analysis from images of natural disasters,” in International conference on image analysis and processing, 2019, pp. 104–113.
J. Chen, Q. Mao, and L. Xue, “Visual sentiment analysis with active learning,” IEEE Access, vol. 8, pp. 185899–185908, 2020, doi: 10.1109/ACCESS.2020.3024948.
Ortis, G. M. Farinella, G. Torrisi, and S. Battiato, “Exploiting objective text description of images for visual sentiment analysis,” Multimed. Tools Appl., vol. 80, no. 15, pp. 22323–22346, 2021, doi: 10.1007/s11042-019-08312-7.