File size: 2,450 Bytes
b2a5e43 85756d9 4ecf744 85756d9 3a811ce b2a5e43 3a811ce 85756d9 963246d bb530fb 963246d 85756d9 bb530fb 85756d9 d038a56 85756d9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 |
---
language:
- en
license: mit
tags:
- text-classification
widget:
- text: "Why do we need an NFQA taxonomy?"
---
# Non Factoid Question Category classification in English
## NFQA model
Repository: [https://github.com/Lurunchik/NF-CATS](https://github.com/Lurunchik/NF-CATS)
Model trained with NFQA dataset. Base model is [roberta-base-squad2](https://huggingface.co/deepset/roberta-base-squad2), a RoBERTa-based model for the task of Question Answering, fine-tuned using the SQuAD2.0 dataset.
Uses `NOT-A-QUESTION`, `FACTOID`, `DEBATE`, `EVIDENCE-BASED`, `INSTRUCTION`, `REASON`, `EXPERIENCE`, `COMPARISON` labels.
## How to use NFQA cat with HuggingFace
##### Load NFQA cat and its tokenizer:
```python
from transformers import AutoTokenizer
from nfqa_model import RobertaNFQAClassification
nfqa_model = RobertaNFQAClassification.from_pretrained("Lurunchik/nf-cats")
nfqa_tokenizer = AutoTokenizer.from_pretrained("deepset/roberta-base-squad2")
```
##### Make prediction using helper function:
```python
def get_nfqa_category_prediction(text):
output = nfqa_model(**nfqa_tokenizer(text, return_tensors="pt"))
index = output.logits.argmax()
return nfqa_model.config.id2label[int(index)]
get_nfqa_category_prediction('how to assign category?')
# result
#'INSTRUCTION'
```
## Demo
You can test the model via [hugginface space](https://huggingface.co/spaces/Lurunchik/nf-cats).
[](https://huggingface.co/spaces/Lurunchik/nf-cats)
## Citation
If you use `NFQA-cats` in your work, please cite [this paper](https://dl.acm.org/doi/10.1145/3477495.3531926)
```
@misc{bolotova2022nfcats,
author = {Bolotova, Valeriia and Blinov, Vladislav and Scholer, Falk and Croft, W. Bruce and Sanderson, Mark},
title = {A Non-Factoid Question-Answering Taxonomy},
year = {2022},
isbn = {9781450387323},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3477495.3531926},
doi = {10.1145/3477495.3531926},
booktitle = {Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval},
pages = {1196–1207},
numpages = {12},
keywords = {question taxonomy, non-factoid question-answering, editorial study, dataset analysis},
location = {Madrid, Spain},
series = {SIGIR '22}
}
```
Enjoy! 🤗 |