The project contains the collected data and code of our paper Yi-Lin Tuan, Yun-Nung Chen, Hung-yi Lee. "DyKgChat: Benchmarking Dialogue Generation Grounding on Dynamic Knowledge Graphs", EMNLP 2019.

The extended abstract version is called Dynamic Knowledge-Grounded Dialogue Generation through Walking on the Graph.

• our proposed approach: (Qadpt) Quick Adaptive Dynamic Knoledge-Grounded Neural Converation Model (pronouce: Q-adapt)

• python3.6
• tensorflow r1.13
• jieba
• nltk3.2.5

• data/: the collected data hgzhz/ and friends/ as well as the trained TransE
• model_ckpts/: the trained models in the paper

• clone the repository
• run the script run.sh

$bash run.sh <GPU_ID> <method> <model> <data> <exp_name>

• for <GPU_ID>, check your device avalibility by nvidia-smi
• for , choose from train, pred_acc, eval_pred_acc, ifchange
• for , choose from seq2seq, MemNet, TAware, KAware, Qadpt
• for , choose from friends, hgzhz_v1_0(used in our paper), hgzhz(current newest version)
• for <exp_name>, check the directory model_ckpts

• testing method
  • pred_acc: for metrics Generated-KW, BLEU-2, distinct-n
  • eval_pred_acc: for metrics KW-Acc, KW/Generic, perplexity
  • ifchange: for change rates / accurate change rates
• script options
  • the hops_num and change_level are required to be changed in run.sh