Understanding Emotive Response to Textual Stimuli: A Multimodal Approach

In an era where human activities are increasingly intertwined with technology, electronic systems and devices are expected to be intelligent and capable of responding in the most human-like way possible.However, for artificial intelligence systems to integrate seamlessly into human society, they must possess emotional intelligence -- the ability to identify, understand, and react to human emotions. As interactions between humans and intelligent systems often rely on language, a deeper comprehension of the link between language and human emotional response becomes crucial for improving human-computer interactions. This thesis aims to understand which aspects of language are connected to emotional response and elicitation. Specifically, it applies an extensive array of linguistically motivated features designed to capture the stylistic elements of language and investigates their relation to human emotions, both in isolation and in combination with other features. Initially, the thesis explores how this broad set of linguistic features correlates with perceived language complexity and the role of language as an emotion elicitor. This same set of features is then used to understand the influence of third-party language on the emotions of a group of subjects. Lastly, this work examines the impact of deliberately emotive language on bodily responses and the relationship between such responses and the previously mentioned linguistic features.