headacheboy / data-of-multimodal-sarcasm-detection Goto Github PK

help

Dear sharer, I haven't found some data about picture attributes in this data set. Is it not provided?

I'm gettting the below exception when executing the main.py script

`
Traceback (most recent call last):

File "main.py", line 181, in
m = Main()
File "main.py", line 22, in init
self.texti = loadData.TextIterator(self.batchSize, self.seqLen)
File "/home/ichimichi/colab/data-of-multimodal-sarcasm-detection/codes/loadData.py", line 25, in init
self.word2id = self.getVocab()
File "/home/ichimichi/colab/data-of-multimodal-sarcasm-detection/codes/loadData.py", line 48, in getVocab
return pickle.load(file)
TypeError: a bytes-like object is required, not 'str'
`

Is there a problem with the dataset_image.zip file? Why does the extraction fail every time?

Hallo, can you share the experimental environment about your code, I cant find it in paper.

Traceback (most recent call last):
File "main.py", line 182, in
m.trainingProcess()
File "main.py", line 160, in trainingProcess
batchX, batchImage, batchWord, batchY = self.texti.nextBatch()
File "/home/wzh/PycharmProjects/data-of-multimodal-sarcasm-detection/codes/loadData.py", line 119, in nextBatch
image = np.load(imagePrefix+str(ID)+".npy")
File "/home/wzh/miniconda3/envs/thomas/lib/python3.6/site-packages/numpy/lib/npyio.py", line 415, in load
fid = open(os_fspath(file), "rb")
FileNotFoundError: [Errno 2] No such file or directory: '../imageVector2/822588646796775425.npy'

Do we need write a ResNet extractor？

Dear sharer, I found that the dataset is missing some pictures. Is it not provided?

We find some images of the training set are missing. Are these images supposed to be missing or something wrong with the uploaded images? If these images are supposed to be missing, how do you process the data without images?

An example of missing images:
910308516510011393
725333760762363905
854334602516733952
904880359350964224
935145326553456641
718073661895249921
851513665983524864
721995610493050880
......

Thanks

Thank you very much for providing the processed data set, but I am very interested in your process of handling emoji, and would like to know what materials you refer to when handling emoji, such as converting 😀 to 'emoji_2655'.

Hi,
We want to analysis the relationship between emoji and sarcasm. However, the form of the emoji code in your dataset is like emoji_XXX. I want to ask if I can know the mapping from the unicode to the emoji_XXX?

Hello, I tried various methods to train vector and vocab in the word folder, but they all failed. At present, Stanford’s glove is the most contacted. How to train vector files in the same format as yours?

Where can I find the original image data?

What's the number in the first column? How can I get the original dataset?(第一列的数字是什么？我怎么能获取原数据集？)

请问一下能在read.me中详细说明本项目的版本（如tensorflow）吗？
如果能这样，非常感激！

When I got 24636 images after decompression, and then I used ResNet-50 V2 extractor to get 24636 .npy files, the shape of every image feature is [1, 14, 14, 2048], why did I still get errors?

Hello:
Thanks for your work, I run the codes and get the attention weight about image，but how can I visulize it, I search a lot, but cant find some useful information.

Dear sharer, thanks a lot for your work., but where can I find the original image data?