Dataset for Research on the Method of Predicting Technical Application Scenarios Based on BERTopic and Recommendation Algorithms

This dataset supports the paper “Research on the Method of Predicting Technical Application Scenarios Based on BERTopic and Recommendation Algorithms.”The paper uses patent data and employs a prediction framework composed of BERTopic-based technical topic extraction, patent knowledge graph construction, and a knowledge-graph-enhanced recommendation module to quantitatively evaluate the potential applications of technologies.The dataset mainly includes:1）Raw data collected from the Derwent patent database, such as patent titles, abstracts, and bibliographic information;2）Results of BERTopic topic analysis;3）Data of the constructed patent knowledge graph;4）Training and test sets built from the co-occurrence relationships between technologies and application scenarios;5）Experimental results and related visualization charts.