KEYWORDS EXTRACTION AND SEMANTIC TREE RESEARCH BASED ON MIXED FEATURES AND LINK INFLUENCE IN LARGE-SCALE DATA

Since the traditional keyword recognition methods cannot effectively combine the semantic and structural information of words, this paper proposes a keyword recognition method, which based on joint feature mining and analysis of word semantic network and co-occurrence structural network. The vocabulary influence network combining the semantic network and the structural network of the text was obtained. Link influence index was proposed to identify keywords. A large-scale Semantic tree of English words was constructed and analyzed by association mining. The experimental results show that the proposed method has a good keyword recognition effect on large-scale corpus data, and the semantic tree obtained by mining can reflect the contextual structure relationship and semantic information of words.