计算机应用与软件2025，Vol.42Issue(5)：271-281,11.DOI:10.3969/j.issn.1000-386x.2025.05.037
基于混合特征和链接影响力的关键词识别及语义树分析
KEYWORDS EXTRACTION AND SEMANTIC TREE RESEARCH BASED ON MIXED FEATURES AND LINK INFLUENCE IN LARGE-SCALE DATA
摘要
Abstract
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.关键词
关键词抽取/图模型/BERT/语义树/影响力Key words
Keyword extraction/Graph model/BERT/Semantic tree/Influence force分类
信息技术与安全科学引用本文复制引用
崔宝阳,冶忠林,赵海兴..基于混合特征和链接影响力的关键词识别及语义树分析[J].计算机应用与软件,2025,42(5):271-281,11.基金项目
国家重点研发计划项目(2020YFC1523300) （2020YFC1523300）
青海省重点研发与转化计划项目(2020-GX-112) （2020-GX-112）
青海省自然科学基金青年项目(2021-ZJ-946Q). （2021-ZJ-946Q）
