Addressing topic modeling with a multi-objective optimization approach based on swarm intelligence
作者：
Highlights：
• Topic Modeling (TM) has been addressed from a multi-objective point of view.
• A new approach based on MOABC has been designed, implemented, and applied to TM.
• Three objective functions have been optimized: coherence, coverage, and perplexity.
• Documents from the Reuters-21578 and TagMyNews datasets have been used for evaluation.
• Results show that MOABC provides relevant improvements with respect to LDA and MOEA/D.
摘要
•Topic Modeling (TM) has been addressed from a multi-objective point of view.•A new approach based on MOABC has been designed, implemented, and applied to TM.•Three objective functions have been optimized: coherence, coverage, and perplexity.•Documents from the Reuters-21578 and TagMyNews datasets have been used for evaluation.•Results show that MOABC provides relevant improvements with respect to LDA and MOEA/D.
论文关键词：Artificial bee colony,Evolutionary computing,Latent Dirichlet allocation,Multi-objective optimization,Text analysis,Topic modeling
论文评审过程：Received 13 January 2021, Revised 2 April 2021, Accepted 2 May 2021, Available online 3 May 2021, Version of Record 6 May 2021.
论文官网地址：https://doi.org/10.1016/j.knosys.2021.107113
