nustm / ecpe Goto Github PK

I wonder how to solve this problem，please！

Hello, I am a graduate student in Computer Science of South China Normal University. Does your team have a PPT for this paper report, and is it convenient for disclosure? I am going to report this paper recently, and I want to use your PPT to make the report better. If you take the liberty, please understand. My email is [email protected] ,if convenient, please send PPT to this email. Thank you !

Hi ! Is the p_cause.py same as p_emotion.py ?

To run the model needs a embedding file named "w2v_200.txt". However, it isn't included in the repository and the description in the article is obscure:
We use word vectors that were pre-trained on the corpora from Chinese Weibo1 with word2vec
(Mikolov et al., 2013) toolkit.
The reference to the website: http://www.aihuang.org/p/challenge.html is invalid. Do you mean this embedding file is trained by yourselves? Otherwise, could you please provide this file in the repository? Thx!

比如all_data_pair.txt 中缺少doc_id编号为9等等一些语料

error: No such file or directory: 'pair_data/P_emotion_1/fold1_train.txt'
when i execute the second step: python pair.py, this problem happened.
if i need to put some file to this folder like the file in the other folder? i can't understand which file to be used.

It's observed that the method run() has been executed twice. Is the twice execution just for getting a reliable result?

Trying to find the English version of the Emotion Cause dataset, but couldn't. Do you have any information of English version dataset?

The paper mentions distance to be used in filtering the candidate emotion cause pairs. What is this distance that has been used ?

The following error was found while running the code
FileNotFoundError: [Errno 2] No such file or directory: '../data/w2v_200.txt'
I finally found data in RTHN, maybe you can upload it.

In addition, could you explain what the files in data_combine are for?

w_cause = tf_funcs.get_weight_varible('softmax_w_cause', [2 * FLAGS.n_hidden, FLAGS.n_class])
b_cause = tf_funcs.get_weight_varible('softmax_b_cause', [FLAGS.n_class])

Where's '/data/w2v_200.txt'

I have a question about lambda parameter.

In your paper, the lambda refers tradeoff parameter applied to loss function.

But in this code, cause_lambda is 1 and also pos_lambda is 1, not 0.

Could you please describe this issue?