Recommendation systems (RS) are widely employed for personalized recommendations across various domains. This study focuses on developing a content generation system utilizing web scraping and automatic text summarization to enhance RS capabilities. To evaluate the quality of generated text, we propose an indicator-based approach that leverages text analysis techniques. The work presents a novel framework for generating recommendations by classifying text based on its relevance to user preferences. The study encompasses two key contributions: 1) the development of a content generation system and 2) the initial implementation of the “Context of the Sources” module within the larger RS framework. The analysis is centered on evaluating the “Context of the Sources” module’s performance across diverse domains (academic, security, non-security), employing text classification as a primary evaluation method. The results provide insights into the efficacy and potential for implementation of this module in real-world applications.