Introduction

Kedma Hamelberg is a PhD Researcher in the Marketing department. In her research project, "Artificial Intelligence for Responsible Marketing: Information Mining from Digital Content”, she investigates B2C online communication and its effects on society. She applies cutting-edge methodologies from Data Science (e.g., Natural Language Processing, Deep Learning) to analyse the responsible use of unstructured data in marketing-customer analytics. More specifically, it involves the following projects:

(Project 1) How do stakeholders react when CEOs do (not) walk the talk? Business implications of CEO socioenvironmental activism.
In this project, she investigates stakeholders' reactions to the (mis)alignment between CEOs' communication on socioenvironmental values and their actions (i.e. climate change and the Russia-Ukraine war).

(Project 2) Brand transcendence and climate change: Leveraging cofollowership patterns on Social Media to identify win-win opportunities for brands and society. 
Climate change is a complex challenge. Will acting together (i.e. strategic alliance between different groups/brands) benefit the cause? With the results from this project, she can guide brands on which alliances would benefit businesses and society based on cofollowership patterns.

(Project 3) How do the various elements of brand linguistic identity impact customer compliance/engagement?
This project identifies the key elements of brand linguistic identity (e.g. textual paralanguage, emotions) and their effects on consumers to enhance the responsible use of brand content on digital platforms. More specifically, we classify textual and visual (i.e., emojis) elements from brand social media posts and their responsible use for increasing customer compliance/engagement.

(Project 4) The two alternatives:
Reading between the lines of online customers' complaints and insatisfactions. On the one hand, brands can use diverse linguistic constructions to ask, "How can I help you?". On the other hand, the narrative structure used by customers can create emphasis or weaken some aspects of the discourse and reveal which brand elements are influencing customers' responses. Thus, the project examines how exactly linguistic elements of brand-customer online conversations (e.g. topics, promises, and emotions) can improve online customer service.

What drives prosocial behaviour?
How to bridge the say-do gap in prosocial behaviour? How do the different linguistic elements from crowdsourcing campaigns influence donation performance (i.e. increase in amount per donation, the total amount of donation, the total number of donations)? By examining online social crowdfunding data of targeted (i.e., online platforms) and non-targeted audiences (i.e., social media), this project intends to uncover the motivations of online donations and how the different elements from a post influence donation performance.