Modeling-the-classification-of-users-into-various-types-of-lifestyles-tribes-based-on-tribal-marketing-in-Twitter

The emergence of web 2.0 and the social media platforms created for different personalities led to the increasing daily involvement of people globally with these platforms. Due to the comprehensive communications through social networks, there are numerous opportunities for individuals, groups, and businesses to analyze these virtual environments.

To enhance the marketing strategies and to identify the most potential customers, we surveyed selected influencers of particular lifestyle tribes such as “Travel, Fitness, Fashion, Art, and Vegan” to train and create a model. This model was based on machine learning techniques, text processing, and web scrapping to present an effective tool to automatically classify Twitter users into pre-defined tribes upon their Twitter pages’ details. We could nearly scrape 45,000 tweets for 420 related influencers and have them as training set in the model. After that, we evaluated the parameter optimization to reach the best parameters for our model. Eventually, we achieved adequate accuracy parameters, an accuracy of 74.3%, and a recall of 71.5%. We intended to get further feedback, so by deploying the model, we selected a sample of 10 reputable brands in each tribe to compare the brands’ mindset with their virtual tribes’ facts.

Keywords: Tribal Marketing, Social Media Analysis, Twitter, Virtual Tribes, Lifestyle, Travel, Fitness, Fashion, Art, Vegan, KNIME Analytics Platform