A Streamlit-Based Interactive Framework for Exploratory Data Analysis, Model Building, Intelligent Recommendations, and Real-Time PubMed Article Search

The task of interpreting and applying medical journal citation data is to be able to explore big data and find meaning while serving relevant advice in real time. To this end, the present study devises an interactive framework based on Streamlit providing exploratory data analysis (EDA), machine learning modeling, a recommendation system, and on a fly article search. This forms a structuring tool of comprehensive analysis through visualizations like word clouds, co-authorship matrice and top authors. It applies a Logistic Regression model trained on the textual data to assess the sentiment and uses cosine similarity + TF-IDF vectorization to create intelligent journal suggestions. Users can ingest real-time PubMed articles using a dedicated module using specific filters such as title keywords and publication year. It shows how these interactive data-driven applications can scale and become usable, bridging the gap between advanced analytical capabilities and making research more accessible via real-world datasets.