Construction and evaluation of ontological tag trees
作者：
Highlights：
• Tag trees are constructed to capture corpus statistics with low space requirements.
• They are constructed using semantic ontologies and corpus statistics.
• The formulated optimization is solved using local search paradigm.
• WordNet is used to solve optimization faster by initializing the local search.
• Completely automated evaluation tasks are proposed to evaluate tag trees.
摘要
•Tag trees are constructed to capture corpus statistics with low space requirements.•They are constructed using semantic ontologies and corpus statistics.•The formulated optimization is solved using local search paradigm.•WordNet is used to solve optimization faster by initializing the local search.•Completely automated evaluation tasks are proposed to evaluate tag trees.
论文关键词：Tag tree,Local search,Optimization,Tag prediction,Classification
论文评审过程：Received 7 January 2015, Revised 23 July 2015, Accepted 24 July 2015, Available online 7 August 2015, Version of Record 27 September 2015.
论文官网地址：https://doi.org/10.1016/j.eswa.2015.07.057
