cuilimeng/deterrent

deterrent's Issues

Training/Testing json data

Thank you for sharing the code. But seems the training json file only contains 1 json data. Is there any guide for generating more for training?

Besides, is it possible to upload a pretrained model? So we can just run through the testing process to get the reasonable results.

Thank you!

Question: What is the Semantic Group information used for the experiment for HGAT[24]? How to obtain this information?

Dear authors,

Thanks a lot for sharing the code!

May I ask a small question regarding the experiments?
In the paper (Appendix), you mention that the HGAT code is implemented by your selves. I refer to Section 5.2 for the details to implement HGAT, where it states:

"HGAT [24]: HGAT is a flexible heterogeneous information network framework for classifying short texts, which can integrate
any type of additional information. We add Semantic Group to the entities as side information, such as Procedures and Disorders."

May I ask what is this Semantic Group information utilized as side information of the entities? And how can we obtain this information?

Thanks a lot for your time!

How to locate the attention weights of the triples in Fig 5 in the paper?

Dear authors,

Thanks a lot for sharing the code of the paper! It really inspires me a lot.
May I ask one small question? In the case study illustrated in Figure 5 in the KDD 2020 paper, there is a attention weight of each triple. May I know whether this attention weight refers to the corresponding value of each entity pair and the j-th relation in line 76 in TextRelationalGraphAttention.py?

A = F.softmax(A, dim=-1)

Or is it the score computed in line 78 in DETERRENT.py?

score = triple_heads * triple_relations * triple_tails

If no, could you please inform me which variables in the code refer to these attention weights?

Thank you so much again for your time! :)

output.csv

Thanks for this study. I have a question about what output.csv clarifies and how it can be created.

Thanks in advance.