Regresar

Model for Enriching Social Media Hashtags Using AI for Institutional Image Management: Case of the Municipality of Ecuador

Abstract:

Governmental entities strive to effectively communicate with the public by disseminating information through social networks. However, the conventional use of static hashtags by community managers hampers the tracking of user opinions and inhibits the acquisition of citizen-enriched hashtags. Furthermore, existing social network keyword analysis services often fall short in domains like municipalities and languages with intricate syntactic and semantic complexities, such as Spanish. This research aims to address these challenges by harnessing artificial intelligence techniques, including Natural Language Processing, Sentiment Analysis, and Topic Modeling, to enhance information dissemination and engagement. The core objective is to augment the visibility of governmental information improving their marketing strategies. This study presents a comprehensive scheme founded on Knowledge Discovery in Databases (KDD). This scheme analyzes Facebook posts within the specialized domain of the Municipality in Ecuador, discerning pertinent keywords and topics through the application of Named Entity Recognition and Latent Dirichlet Allocation (LDA). The data processing methodology adopts an Extraction, Transform, and Load (ETL) framework, ensuring the generation of high-quality data for in-depth analysis. The culmination of this research is realized through the development of an intuitive dashboard application, following the Model-View-Controller (MVC) architecture. This application serves as a valuable tool for data visualization, significantly advancing the automation of data analysis in governmental social networks. Ultimately, this research stands to benefit governmental organizations by facilitating more effective interaction with their target audience and offering insights for the enhancement of communication and marketing strategies.