An Educational Knowledge Graph Question Answering Dataset with Cognitive Expressions and Language Styles

Educational question answering is vital for online education, with Educational Knowledge Graph Question Answering (EKGQA) focusing on leveraging knowledge bases to answer questions. However, the lack of real Chinese annotated datasets hampers EKGQA's progress. Current datasets often feature direct questions, but actual educational questions reflect varied cognitive expressions and language styles due to differing experiences and backgrounds. To address this, in this paper, we propose to benchmark such challenges for EKGQA by establishing a more realistic dataset EDUCEQ. It contains 236.3k questions in 12 categories. Compared with existing datasets, EDUCEQ features two more authentic features - diverse cognitive expressions and various language styles.Additionally, we have evaluated various categories using the EDUCEQ dataset and developed a robust baseline EKGQA method that establishes competitive benchmarks for future research.The EDUCEQ is the first to portray a more realistic view of educational phenomena from a cognitive and computational linguistics perspective, which is beneficial to educational knowledge graph question answering and also the online education field in general.