Source code and dataset for our paper "Difficulty-controllable Multi-hop Question Generation From Knowledge Graphs" accepted at ISWC 2019.
- Download pre-trained embeddings from here (https://139.129.163.161/index/toolkits#pretrained-freebase)
- Preprocess cwq dataset and create binary file data.pt (which also contains a vocabulary)
- Extract embeddings for entities and relationships in CWQ dataset (given in this repo) from vocabulary.
- Create an embedding file with all entities, relationships and 256 dimensional embeddings (all space separated) say pretrained.pt
- Run train.py with '-data' path set to data.pt use this pretrained embedding with -usepretrained argument
- To generate a question from the trained model run "generate_question.py" with required arguments such as model path etc.