Querying knowledge graphs in natural language

Abstract Knowledge graphs are a powerful concept for querying large amounts of data. These knowledge typically enormous and often not easily accessible to end-users because they require specialized in query languages such as SPARQL. Moreover, need deep understanding the structure underlying data models based on Resource Description Framework (RDF). This drawback has led development Question-Answering (QA) systems that enable express their information needs natural language. While existing simplify user access, there is still room improvement accuracy these systems. In this paper we propose new QA system translating language questions into SPARQL queries. The key idea break up translation process 5 smaller, more manageable sub-tasks use ensemble machine learning methods well Tree-LSTM-based neural network automatically learn translate question query. performance our proposed empirically evaluated using two renowned benchmarks-the 7th Question Answering over Linked Data Challenge (QALD-7) Large-Scale Complex Dataset (LC-QuAD). Experimental results show outperforms state-of-art by 15% QALD-7 dataset 48% LC-QuAD dataset, respectively. addition, make source code available.