This website requires JavaScript.

Author Name Disambiguation via Heterogeneous Network Embedding from Structural and Semantic Perspectives

Wenjin XieSiyuan LiuXiaomeng WangTao Jia
Dec 2022
摘要
Name ambiguity is common in academic digital libraries, such as multipleauthors having the same name. This creates challenges for academic datamanagement and analysis, thus name disambiguation becomes necessary. Theprocedure of name disambiguation is to divide publications with the same nameinto different groups, each group belonging to a unique author. A large amountof attribute information in publications makes traditional methods fall intothe quagmire of feature selection. These methods always select attributesartificially and equally, which usually causes a negative impact on accuracy.The proposed method is mainly based on representation learning forheterogeneous networks and clustering and exploits the self-attentiontechnology to solve the problem. The presentation of publications is asynthesis of structural and semantic representations. The structuralrepresentation is obtained by meta-path-based sampling and a skip-gram-basedembedding method, and meta-path level attention is introduced to automaticallylearn the weight of each feature. The semantic representation is generatedusing NLP tools. Our proposal performs better in terms of name disambiguationaccuracy compared with baselines and the ablation experiments demonstrate theimprovement by feature selection and the meta-path level attention in ourmethod. The experimental results show the superiority of our new method forcapturing the most attributes from publications and reducing the impact ofredundant information.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?