Dataset for \u201cA Data-driven Framework for Identifying Key Technologies and Analyzing Evolutionary Patterns\u201d

This dataset includes large-scale patent citation and semantic similarity data used in the study \u201cA Data-driven Framework for Identifying Key Technologies and Analyzing Evolutionary Patterns.\u201d It contains 940,369 AI-related patents and 1,709,522 citation relations, along with Sentence-BERT embeddings and preprocessed metadata for graph construction.