Complex Knowledge Base Question Answering: A Survey

Knowledge base question answering (KBQA) aims to answer a over knowledge (KB). Early studies mainly focused on simple questions KBs and achieved great success. However, their performances complex are still far from satisfactory. Therefore, in recent years, researchers propose large number of novel methods, which looked into the challenges questions. In this survey, we review advances KBQA with focus solving questions, usually contain multiple subjects, express compound relations, or involve numerical operations. detail, begin introducing task relevant background. Then, present two mainstream categories methods for KBQA, namely semantic parsing-based (SP-based) information retrieval-based (IR-based) methods. Specifically, illustrate procedures flow designs discuss difference similarity. Next, summarize that these encounter when explicate advanced solutions as well techniques used existing work. After that, potential impact pre-trained language models (PLMs) KBQA. To help readers catch up SOTA also provide comprehensive evaluation resource about task. Finally, conclude several promising directions related future research.