CKBQA Task for Chinese Knowledge Graph Question Answering

This content describes the CKBQA (Chinese Knowledge Base Question Answering) task introduced at CCKS2021, focusing on natural language question answering using Meituan's life services domain knowledge graph. It outlines the task's objectives, including handling both domain-specific and open-domain questions, and emphasizes the need for systems to process complex natural language queries accurately. The task integrates structured knowledge graphs with open-domain data from PKUBASE, aiming to advance KBQA research and applications in practical scenarios like smart customer service and complex search.

The evaluation criteria prioritize accuracy, efficiency, and explainability of KBQA systems. The competition offers substantial rewards, encouraging innovations in semantic parsing, information retrieval, and domain adaptation. The task also highlights the growing demand for KBQA in real-world applications such as intelligent voice assistants and e-commerce platforms.