Topic Selection in Latent Dirichlet Allocation

摘要：

　　Latent Dirichlet Allocation(LDA)has been widely applied to text mining.LDA is a probabilistic topic model which processes documents as the probability distribution of topics.One challenging issue in application of LDA is to select the optimal number of topics in LDA model.This paper presents a topic selection method which considers the density of each topic and computes the most unstable topic structure through an iteration process.Evaluation results show that the proposed method can generate an optimal number of topics automatically with a small number of iterations.

关键词： MapReduce job scheduling data locality

作者: Biao Wang Yang Liu Zelong Liu Maozhen Li Man Qi

作者单位: State Grid Sichuan Electric Power Research Institute Chengdu, China School of Electrical Engineering and Information Systems Sichuan University Chengdu, 610065, China School of Engineering and Design Brunel University Uxbridge, UB8 3PH, UK Department of Computing Canterbury Christ Church University Canterbury, Kent, CT1 1QU, UK

会议类型: 国际会议

会议名称: The 2014 10th International Conference on Natural Computation (ICNC 2014) and the 2014 11th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD 2014)(第十届自然计算和第十一届模糊系统与知识发现国际会议)

会议地点: 厦门

会议语种:英文

页码: 766-770

在线出版日期: 2014-08-19（万方平台首次上网日期，不代表论文的发表时间）