What dose "positive_head" mean in line 197 in "KGQA/LSTM/main.py"

[Ironeie]

for i_batch, a in enumerate(loader): model.zero_grad() question = a[0].to(device) sent_len = a[1].to(device) positive_head = a[2].to(device) positive_tail = a[3].to(device) Since question represents the embedding of questions, sent_len represents the length of questions, positive_tail represents the lables(answers) of questions, what dose "positive_head" represent?

[apoorvumang]

positive_head is the entity present in the question. As followed by others, we assume entity linking to be 'perfect' ie. the model knows exactly which entity is there in the question. There is always exactly one entity.

[ShuangNYU]

positive_head is the entity present in the question. As followed by others, we assume entity linking to be 'perfect' ie. the model knows exactly which entity is there in the question. There is always exactly one entity.

Is that true that positive_tail is a one-hot vector but only one position is '1'?

[Ironeie]

positive_head is the entity present in the question. As followed by others, we assume entity linking to be 'perfect' ie. the model knows exactly which entity is there in the question. There is always exactly one entity.

Is that true that positive_tail is a one-hot vector but only one position is '1'?

I tried to print the positive_tail and found that there are more than one position being '1', where I don't understand why it is.

[apoorvumang]

positive_head is the entity present in the question. As followed by others, we assume entity linking to be 'perfect' ie. the model knows exactly which entity is there in the question. There is always exactly one entity.

Is that true that positive_tail is a one-hot vector but only one position is '1'?

I tried to print the positive_tail and found that there are more than one position being '1', where I don't understand why it is.

Multiple answers can be correct, so it's k-hot not 1-hot