Mapping trend travel trajectory of Australian food with Instagram influencer network map

• Main Author: Yan, Kaizhen
• Other Authors: Lihui CHEN
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::General; Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/149310
• Institution: Nanyang Technological University

Social media data has emerged to be the best source for brands to gain insights on consumers’ behaviour and preferences. This creates a vast amount of commercial value as brands can track the evolution of consumer preferences and update their marketing or product strategies accordingly. The existing approaches to analyse social media, however, fail to exploit the potential value of this consumer data. On one hand, methods like keyword sizing or sentiment detection only capture the mainstream conversation but cannot shed light on the next potential opportunity. On the other hand, trend detection frameworks like topic modeling only detect trends, but lack context, making it difficult to assist brands making holistic decisions. This project proposes a new framework for detecting and tracking consumer generated trends. The output of this project is also expected to be integrated into Synthesis’ consulting projects that are relevant to trend detection. The proposed framework is mainly based on network theory, including community detection, authority computation etc. A proof of concept will be provided to validate this new analysis framework. The network maps of food influencers on Instagram is created for the experimental study. We’ll use these maps to investigate how food trends spread among the crowd, who are the main drivers and how these insights could potentially bring value to consumer brands. The report consists of three major parts. First part focuses on the literature review and selection of research and the methodology to come up with a new trend analysis framework. Second part elaborates the actual implementation and the experimental study on the Instagram Australian food influencers dataset. Third part discusses the business value and implication of this project output.