基于加权超图随机游走的文献关键词提取算法
|
摘要
针对科技文献类标题短文本关键词提取时,已有自然语言处理算法难以建模文献时间与权威性且短文本词语较少建模往往存在高维稀疏问题,本文提出了一个综合实时性以及权威性的关键词提取算法为研究者进行相关推荐.该方法将文献标题视为超边,将标题中不同词项视为超点来构建超图,并对超图中的超边与超点同时加权,进而设计一种基于加权超图随机游走的关键词提取算法对文献标题的词项进行提取.该模型通过对文献来源,发表年份以及被引次数建模来对超边进行加权,根据节点之间的关联度以及每对节点在特定标题中的共现距离对超点加权.最后,通过超图上的随机游走计算出节点的重要性进而确立可推荐的关键词.实验表明,与三种基准短文本关键词提取算法相比,本文算法在精确率和召回率方面均有所提高.
It is difficult for the existing natural language processing algorithms to model the time and authority of short texts such as paper titles of scientific literature. Besides,the short texts always tend to have fewer words and thus suffer from high dimension and sparsity. A keyword extraction method involving both real-time and authoritativeness is presented.A weighted hyper-graph is constructed where vertexes represent weighted terms and weighted hyper-edges measure the semantic relatedness of both binary relations and nary relations among terms. On one hand,the source of the documents,the year of publication and number of citations are considered for weighting hyper-edges,on the other hand,the degree of association between the nodes and co-occurrence distance for each pair of nodes in particular title are calculated for weighting hypervertexes. The random walk approach is performed on the weighted hyper-graph to obtain the recommended keywords.Experimental results demonstrated that compared with three baseline algorithms,the proposed approach is able to extract keywords with higher precision and recall.
It is difficult for the existing natural language processing algorithms to model the time and authority of short texts such as paper titles of scientific literature. Besides,the short texts always tend to have fewer words and thus suffer from high dimension and sparsity. A keyword extraction method involving both real-time and authoritativeness is presented.A weighted hyper-graph is constructed where vertexes represent weighted terms and weighted hyper-edges measure the semantic relatedness of both binary relations and nary relations among terms. On one hand,the source of the documents,the year of publication and number of citations are considered for weighting hyper-edges,on the other hand,the degree of association between the nodes and co-occurrence distance for each pair of nodes in particular title are calculated for weighting hypervertexes. The random walk approach is performed on the weighted hyper-graph to obtain the recommended keywords.Experimental results demonstrated that compared with three baseline algorithms,the proposed approach is able to extract keywords with higher precision and recall.
引文
[1]Ma Hui-fang,Xing Yu-ying,et al.Leveraging term co-occurrence distance and strong classification features for short text feature selection[A].Proceedings of the 10th International Conference on Know ledge Science,Engineering and Management[C].Melbourne,Australia:Springer,2017.303-310.
[2]刘喜平,万常选等.空间关键词搜索研究综述[J].软件学报,2016,27(2):329-347.Liu Xi-ping,Wan Chang-xuan,et al.Survey on spatial keyw ord search[J].Journal of Softw are,2016,27(2):329-347.(in Chinese)
[3]Ugo Erra,Sabrina Senatore,et al.Approximate TF-IDF based on topic extraction from massive message stream using the GPU[J].Information Sciences,2015,292(20):143-161.
[4]Song Shao-xu,Zhu Han,et al,Probabilistic correlationbased similarity measure on text records[J].Information Sciences,2014,289(1):8-24.
[5]Hua Wen,Wang Zhong-yuan,et al.Short text understanding through lexical-semantic analysis[A].Proceedings of the 31st International Conference on Data Engineering[C].Seoul,South Korea:IEEE,2015.495-506.
[6]Blei D M,Ng A Y,Jordan MI.Latent Dirichlet allocation[J].Journal of Machine Learning Research,2003,3(1):993-1022.
[7]Rahim Saeidi,Ramon Fernandez Astudillo,et al.Uncertain LDA:Including observation uncertainties in discriminative transforms[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(7):1479-1488.
[8]Willyan D Abilhoa,Leandro N de Castro.A keyword extraction method from tw itter messages represented as graphs[J].Applied Mathematics and Computation,2014,240(4):308-325.
[9]Zhou Deng-yong,Huang Jia-yuan,et al.Learning with hypergraphs:clustering,classification,and embedding[A].Proceedings of the 20th International Conference on Neural Information Processing Systems[C].Vancouver,Canada:MIT Press,2006.1601-1608.
[10]Li De-cong,Li Su-jian.Hypergraph-based inductive learning for generating implicit key phrases[A].Proceedings of the 20th International Conference on World Wide Web[C].Hyderabad,India:Springer,2011.77-78.
[11]Bellaachia A,Al-Dhelaan M.HG-Rank:A hypergraphbased keyphrase extraction for short documents in dynamic genre[A].Proceedings of the 4th Workshop on Making Sense of Microposts[C].Seoul,Korea:CEURWS,2014.42-49.
[2]刘喜平,万常选等.空间关键词搜索研究综述[J].软件学报,2016,27(2):329-347.Liu Xi-ping,Wan Chang-xuan,et al.Survey on spatial keyw ord search[J].Journal of Softw are,2016,27(2):329-347.(in Chinese)
[3]Ugo Erra,Sabrina Senatore,et al.Approximate TF-IDF based on topic extraction from massive message stream using the GPU[J].Information Sciences,2015,292(20):143-161.
[4]Song Shao-xu,Zhu Han,et al,Probabilistic correlationbased similarity measure on text records[J].Information Sciences,2014,289(1):8-24.
[5]Hua Wen,Wang Zhong-yuan,et al.Short text understanding through lexical-semantic analysis[A].Proceedings of the 31st International Conference on Data Engineering[C].Seoul,South Korea:IEEE,2015.495-506.
[6]Blei D M,Ng A Y,Jordan MI.Latent Dirichlet allocation[J].Journal of Machine Learning Research,2003,3(1):993-1022.
[7]Rahim Saeidi,Ramon Fernandez Astudillo,et al.Uncertain LDA:Including observation uncertainties in discriminative transforms[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(7):1479-1488.
[8]Willyan D Abilhoa,Leandro N de Castro.A keyword extraction method from tw itter messages represented as graphs[J].Applied Mathematics and Computation,2014,240(4):308-325.
[9]Zhou Deng-yong,Huang Jia-yuan,et al.Learning with hypergraphs:clustering,classification,and embedding[A].Proceedings of the 20th International Conference on Neural Information Processing Systems[C].Vancouver,Canada:MIT Press,2006.1601-1608.
[10]Li De-cong,Li Su-jian.Hypergraph-based inductive learning for generating implicit key phrases[A].Proceedings of the 20th International Conference on World Wide Web[C].Hyderabad,India:Springer,2011.77-78.
[11]Bellaachia A,Al-Dhelaan M.HG-Rank:A hypergraphbased keyphrase extraction for short documents in dynamic genre[A].Proceedings of the 4th Workshop on Making Sense of Microposts[C].Seoul,Korea:CEURWS,2014.42-49.