Investigating the Value of Sports Footwear Brands using Natural Language Processing Methods
Investigating the Value of Sports Footwear Brands using Natural Language Processing Methods

Author(s): Dijana Vuković, Sara Slamić Tarade
Subject(s): Social Sciences
Published by: Udruženje ekonomista i menadžera Balkana
Keywords: Brand significance; Brand value; Natural Language Processing; Topic modeling; Semantic Brand Score
Summary/Abstract: Determining the value of brands and comparing them by analyzing key elements such as identity, image and value is important for marketing measures and the branding of products or services. This paper presents research findings based on innovative methods for determining the significance and value of sports footwear brands. Natural Language Processing (NLP) techniques are used to analyze extensive text content collected from sports footwear-related websites. To determine the most relevant sports footwear brands, NLP techniques are used for topic modeling based on the Latent Dirichlet Allocation (LDA) method. LDA is an unsupervised method used to determine the topics addressed in the analyzed texts by extracting the most significant words in these topics. The importance of these identified brands in the text corpus is determined using the Semantic Brand Scores method, which uses graph theory to determine the importance of the brand in the text corpus based on three dimensions: prevalence, diversity and connectivity.

• Details
• Contents