-
Notifications
You must be signed in to change notification settings - Fork 15
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use Sarcasm v2 Dataset #1
Comments
v2 is a single csv file. I can write a python function to covert that file into the format learn.datasets.load_files expects? For example, for the following data point: Corpus,Label,ID,Quote Programmatically
This way the current data loading can work as it is. What do you think about this approach? |
v2 Dataset has columns Do not go down the method you proposed. |
It sounds like you are describing a more substantial change. Then what are the steps of achieving what you propose? Since you label this as hacktoberfest, could you provide some more direction? |
Can I work on this issue? What exactly are the problems or concerns regarding this issue at the moment? |
Current Method - We generate features from a single text field to train the models. The desired Method - v2 Dataset provides 2 text field - question and reply to it. We want to make new models based on these 2 inputs. Hope this helps |
Basically this means we have to start from the ground up - we now have to train a model for the replies too, if I'm not wrong? (I'll study the code and see how you trained the first time around.) Plan of action:
|
Could you guide me as to how you created the dataset? |
@Bhanu1911 What I was thinking is little different.
This makes sense because Sarcasm is context based. Having comment and its parent comment will be accurate than a single comment. |
I think the source gives enough background about how they created the dataset - Sarcasm v2 |
I meant how did you partition the dataset? |
Sarcasm v2 is a better dataset for this project. Since it has both parent comment and reply. Apply this dataset to make the prediction better.
The text was updated successfully, but these errors were encountered: