Exploring the sustainable packaging sector through European companies' LinkedIn profiles with unsupervised machine learning.

The rapidly increasing demand for goods and services has led to increased use of packaging for products. Production and consumption practices are critical for environmental preservation as they directly impact pollutants, emissions, and waste generation. This study investigates 516 LinkedIn profiles of entities registered as "companies" that include the term "Sustainable Packaging" in any section of their profiles and whose location filter is set to "Europe". Two unsupervised machine learning approaches, HDBSCAN and Latent Dirichlet Allocation (LDA), are employed to analyse a dataset of company descriptions focusing on the sustainable packaging sector. The first topic emphasizes sustainability across various business operations and product development aspects. The second topic shares many core terms with the first, such as "packaging," "sustainable," and "product," indicating a general thematic overlap. However, it highlights how companies leverage sustainability as a key component of their brand identity, particularly in markets like cosmetics and organic products.