Abstract:
The advancement of the world wide web from its infancy of Web1.0 with read-only web pages allowing minimal interaction to the Web 2.0 which brough more a collaborative era for the internet, I have observed the emergence of social media and other media presence such as blogs allowing for more user created content on subjects of interest. This has allowed many to express their thought and opinions on a multitude of subject which in turn consumed by other Web 2.0 users can influence their decision-making processes. Such shared opinions can be mined to determine an overall sentiment towards a product, service, political climate etc. Such sentiment analysis has mainly focused on textual sentiment mining and well established commercial and open-source applications are available for this purpose. However, limited research has been carried out to directly mine the sentiment of an image. This research focused on the study of determining the current gap in the research of visual sentiment analysis with the aim of developing a model that can be used directly to determine the visual sentiment.
