SParseQA: Sequential word reordering and parsing for answering complex natural language questions over knowledge graphs
作者：
Highlights：
• A sequential word reordering & parsing method builds the interpretations of NLQs.
• The question graph is constructed and refined considering the query graph pattern.
• The ordered dependency tree is leveraged to provide a more stable word ordering.
• A graph similarity process is proposed that exploits a relation pattern taxonomy.
• Extensive experiments reveal significant progress in answering the complex NLQs.
摘要
•A sequential word reordering & parsing method builds the interpretations of NLQs.•The question graph is constructed and refined considering the query graph pattern.•The ordered dependency tree is leveraged to provide a more stable word ordering.•A graph similarity process is proposed that exploits a relation pattern taxonomy.•Extensive experiments reveal significant progress in answering the complex NLQs.
论文关键词：RDF complex question answering,Uncertain question graph construction,Sequential word reordering and parsing,Relation pattern-based similarity
论文评审过程：Received 26 February 2021, Revised 16 October 2021, Accepted 19 October 2021, Available online 22 October 2021, Version of Record 2 November 2021.
论文官网地址：https://doi.org/10.1016/j.knosys.2021.107626
