Social media sites provide warning signs for shifts in consumer behavior, competitive forces, and emerging market trends. However, most small and medium-sized businesses (SMBs) do not have a systematic and scalable approach to tap into this unstructured data from various sites to extract insights. This paper proposes a trend detection method that leverages automated data extraction with n8n workflows, transformer-based embeddings, hybrid sentiment analysis, BERTopic clustering, and a weighted TrendScore composite score. The proposed approach combines multiple, heterogeneous inputs into a single analytical workflow and offers explainable visual and conversational BI interfaces, which are specifically designed for SMBs. The parameter definitions, scoring rules, and workflow diagrams are carefully detailed to ensure that the approach is fully reproducible. The proposed approach focuses on interpretability, robustness across multiple platforms, and applicability within a resource-constrained business setting. • Reproducible multi-platform social data acquisition using exportable n8n workflows. • Hybrid Transformer-Lexicon sentiment modeling combined with BERTopic clustering. • Quantified TrendScore integrating growth, engagement, sentiment shift, and cross-platform consistency.
Baviskar et al. (Sun,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: