Image-based memes as sentiment predictors

Abstract

This paper explores the correlation between the implied semantic meaning of image-based memes and the textual content of discussions in social media. The behavior of a meme is also distinctive as memes replicate and mutate, similar to genes in human evolution, during propagation in social media networks. This research focuses on the use of sentiment analysis to discover affective meanings of textual comments in social media that are emphasized by the use of memes. Social media provides an interesting challenge in the field of semantic textual analysis due to the prevalence of slang within casual communications. The research supports that the affective nature of a topic in social media can be emphasized by the type of meme applied to the communication. It is the relationship between the unusual multimedia element — the meme — and social media textual communications at motivate the research. In addition to identifying correlations between text and memes, trends in the use of popular memes are also identified. This research successfully demonstrated that memes are used to emphasize the semantic content in social media communications and that the meanings of memes correlate to the topics of the discussion threads.