Integración de Inteligencia Artificial en la Inteligencia de Negocio: potencial de los datos no estructurados para la toma de decisiones estratégicas

The integration of Artificial Intelligence (AI) with Business Intelligence (BI) systems is changing the way companies manage and analyze large volumes of data, especially those that do not follow a structured format, such as text, images, and videos. Traditionally, BI has worked with structured data, which is easy to organize and analyze. However, unstructured data, coming from social networks, emails, and other media, is becoming increasingly important, leading companies to implement advanced technologies to take advantage of this information. This research examines how the use of Multiple Regression Models can help combine structured and unstructured data to improve predictions and decision making. The Multiple Regression Model allows analyzing the relationship between several independent variables, both structured (such as sales or prices) and unstructured (such as sentiments extracted from social networks), and a dependent outcome, such as customer satisfaction or future demand. By including unstructured data in the analysis, regression models reveal hidden patterns and relationships that would otherwise be difficult to identify with traditional tools. The results indicate that the use of AI, through techniques such as natural language processing (NLP) and computer vision, significantly improves the accuracy of analyses and predictions when working with large volumes of unstructured data. The ability to process and analyze both structured and unstructured data gives companies a richer and more detailed view of their operating environment, allowing them to adjust their strategies more effectively. However, the research also reveals important limitations. The quality of unstructured data and the ability of companies to manage large volumes of complex information are significant challenges. In addition, the implementation of AI and advanced analysis models, such as multiple regression, requires a robust technological infrastructure and specialized personnel, which can be an obstacle for smaller or resource-limited companies. Ethical and privacy issues also arise that must be addressed, especially when handling sensitive customer data.