NLP - Driven Knowledge Extraction for strategic Decision Making in Management Operations

Abstract

This paper discusses the application of Hybrid NLP with Multimodal Learning (BERT + Vision Transformers) to extract knowledge in strategic decision-making in the management operations. The work uses Hugging Face Transformers Library to incorporate the sophisticated Natural Language Processing (NLP) and Vision Transformers (ViTs), allowing one to analyze both textual and visual data simultaneously. The proposed methodology is a complete system to extract actionable insights as it uses BERT to analyze sentiment and extract entities and ViTs to analyze visual data like product images and customer feedback videos. The findings demonstrate that the accuracy of decision making and trend prediction has a high level of improvement and the hybrid method is better than the traditional NLP models and other multimodal approaches. The approach has issues with regard to computational requirements, and integration with the legacy systems despite performing well. The research will add value to the advancement of stronger data-driven decision support systems in management operations, with the potential of multimodal NLP applications.