LDA-based online topic detection using tensor factorization

• Published in: Journal of information science Vol. 39; no. 4; pp. 459 - 469
• Main Authors: Guo, Xin, Xiang, Yang, Chen, Qian, Huang, Zhenhua, Hao, Yongtao
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 01.08.2013; Sage Publications; Bowker-Saur Ltd
• Subjects: Artificial intelligence; Detection; Dirichlet problem; Exact sciences and technology; Factorization; Feature extraction; General aspects; Information and communication sciences; Information retrieval; Information science. Documentation; Mathematical analysis; Methods; On-line systems; Online; Optimization; Optimization techniques; Sciences and techniques of general use; Simulation; Studies; Subject fields; Temporal logic; Tensors; Texts; Time factors; LDA; topic detection; tensor factorization; topic tensor; On line; Use; Detection; Topic
• ISSN: 0165-5515; 1741-6485
• DOI: 10.1177/0165551512473066

In the information retrieval field, effective and efficient extraction of topics from large-scale online text streams is challenging because it is a fully unsupervised learning task without prior knowledge. Most previous studies have focused on how to analyse text corpus to extract topics, rarely considering time dimensions. In the present study, we approached topic detection as a temporal optimization problem. Here, we propose a novel approach to incremental topic detection, called online topic detection using tensor factorization (OTD-TF), which is based on latent Dirichlet allocation (LDA). First, topics are obtained from the corpus in current time slices using LDA. Second, a topic tensor with a time dimension is constructed to identify the correlations between pairs of topics. Then, approximate topics are merged using TF. Finally, documents are reallocated to corresponding topic bins. By executing these steps continuously and incrementally, temporal topic detection can be achieved. In theoretical analyses and simulation experiments, OTD-TF outperformed other systems in terms of space and time complexity and achieved a high precision ratio. Our experimental evaluations also revealed interesting temporal patterns in topic emergence, development, extinction, burst and transience.