Sažetak
The research objective of this paper is to determine the characteristics of sports shoe models made by two major competing brands, Adidas and Nike. The focus is on selected collections of men’s and women’s sneakers from both brands, based on a data set that includes model names, feature descriptions, average consumer ratings, and the corresponding number of consumer reviews. The textual feature descriptions are analyzed using Natural Language Processing (NLP) techniques such as topic modeling and TF-IDF frequency analysis. Additionally, sentiment analysis, based on the positive and negative consumer ratings and descriptions of sports shoe characteristics, will be conducted to gain insights into user satisfaction.
The topic modeling technique is based on the unsupervised latent Dirichlet method (LDA), which enables the identification of latent topics in text content. In the context of sports footwear, this method can uncover key themes such as design aspects, material composition or performance characteristics, enabling a nuanced understanding of the features highlighted in the descriptions. A quantitative statistical analysis is conducted to test hypotheses about possible differences in the ratings of men’s and women’s sneakers both for individual brands and between them. This research contributes to a deeper understanding of the competitive landscape of the Adidas and Nike
brands in the sports footwear industry and provides valuable insights that can be used to improve marketing strategies.
Ključne riječi
characteristic of footwear, NLP analysis, marketing strategies, sport brands, consumer opinions