Skip to content

chiyuzhang94/PTSM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 

Repository files navigation

Best Paper Award by 1st Workshop on Novel Evaluation Approaches for Text Classification Systems on Social Media @ ICWSM-2022

Datasets

To download PTSM data, please fill out this registration form. We will share the dataset after the registration.

Paraphrase Model

We fine-tune T5-base on four paraphrase gold datasets as shown below.

Dataset Domain # of samples
PIT-2015 Tweet 3,789
LanguageNet Tweet 12,988
Opusparcus Video subtitle 462,846
QQP Quora 149,263
Total - 625,097

You can find our model on: https://huggingface.co/UBC-NLP/ptsm_t5_paraphraser You can load this model by:

from transformers import T5ForConditionalGeneration,T5TokenizerFast

model = T5ForConditionalGeneration.from_pretrained(model_name_or_path)
tokenizer = T5TokenizerFast.from_pretrained(model_name_or_path)

A prefix "paraphrase: " should be added in font of the input sequence, i.e.:

input_st = "paraphrase: " + text + " </s>"

Paraphrase Examples

Original Tweet Paraphrase Gold Label
USER but what I am doing is in my control, #AvoidMMT , you guys are #terrible 1. You guys are horrible, avoid MMT
2. what I am doing is in my control, #AvoidMMT, you guys are terrifying
3. You guys are #terrorist. I have used everything I have to do.
4. You guys are awful, but I am going to stop doing it.
anger
I hate when people say 'I need to talk to you or we need to talk.' My anxiety immediately goes up... 1. I hate when people say 'I need to talk to you or we need to talk to you'. I guess that's the problem.
2. I hate when people tell me 'I need to talk to you or we need to speak' my anxiety immediately goes up.
3. Why am I afraid when people say 'I need to talk to you or we need to talk?'
anger
This is the 47th wedding I've ruined. When I hit 50 I can retire. 1. The 46th wedding I've ruined. When I hit 50 I can retire. It's nice to see yo
2. Here's the 47th wedding I've ruined. If I'm old enough to go on the 40s I can get married.
3. After a single wedding, I drew 47 weddings, and before I hit 50 I can retire"
humor
Sorry guys I have absolutely no idea what time i'll be on cam tomorrow but will keep you posted. 1. Sorry to disturb you. I have absolutely no idea what time I'll be on cam tomorrow.
2. I have absolutely no idea what time I'll be on camera tomorrow but I'll keep you posted
sadness
Who wants to go with me to get my tattoo tomorrow? I'll buy you Dunkin doughnuts 1. "I'll buy you Dunkin' Donuts for $5.
2. Who wants to go with me for my tattoo tomorrow? I'll buy you a Dunkin' Donuts.
3. Who wants to go with me to get my tattoo tomorrow?
neutral
USER May I suggest, that you have a meal that is made with beans, onions & garlic, the day before class. 1. The day before class please eat beans, onions and garlic. Also see the videos
2. "The Day Before Class. You should make that meal, (do you think).
3. If you can eat just the day before class, make a wonderful meal with garlic, onions and beans.
joy

Please find more details in our paper:

@inproceedings{zhang2022decay,
   title={Decay No More: A Persistent Twitter Dataset for Learning Social Meaning},
   author={Zhang, Chiyu and Abdul-Mageed, Muhammad and Nagoudi, El Moatez Billah},
   booktitle ={Proceedings of 1st Workshop on Novel Evaluation Approaches for Text Classification Systems on Social Media (NEATCLasS)}, 
   year={2022},
   url = {https://arxiv.org/pdf/2204.04611.pdf},
   publisher = {{AAAI} Press}, 
}

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published