Existing question answering systems struggle to answer factoid questions when geospatial information is involved. This is because most systems cannot accurately detect the geospatial semantic elements from the natural language questions, or capture the semantic relationships between those elements. In this paper, we propose a geospatial semantic encoding schema and a semantic graph representation which captures the semantic relations and dependencies in geospatial questions. We demonstrate that our proposed graph representation approach aids in the translation from natural language to a formal, executable expression in a query language. To decrease the need for people to provide explanatory information as part of their question and make the translation fully automatic, we treat the semantic encoding of the question as a sequential tagging task, and the graph generation of the query as a semantic dependency parsing task. We apply neural network approaches to automatically encode the geospatial questions into spatial semantic graph representations. Compared with current template-based approaches, our method generalises to a broader range of questions, including those with complex syntax and semantics. Our proposed approach achieves better results on GeoData201 than existing methods.

• python 3.6
• allennlp For LSTM-based semantic encoder.
• transformers For BERT semantic encoder.
• am-parser For graph generation.
• rule-based geo-question encoder To replicate rule-based semantic encoding result.

All data mentioned in the paper is in data directory.

Rule-based encoder: To replicate the result of rule-based encoder, you need to follow the instruction of this to download the relevant tools and gazetteer, then run the simple encoder.

LSTM encoder: We use the allennlp tools, you need to download the GloVe embedding files and ELMo weights, replace the locations in jsonnet file, and run bash lstm_tagger_train.sh.

BERT encoder: We use the huggingface/transformers token-level encoder, simply run bash bert_tagger_train_pred.sh to use the BERT encoder.

Use am-parser to generate DM graph representation, then use graph_build_and_eval.ipynb to generate the graph use the tagging results and DM graph.