Spaces:
Runtime error
Runtime error
File size: 1,409 Bytes
b79be81 7daaa6b b79be81 7daaa6b b79be81 7daaa6b b79be81 7daaa6b 5d51b5a 7daaa6b |
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 |
---
title: Perin
emoji: 📊
colorFrom: purple
colorTo: green
sdk: gradio
sdk_version: 3.1.7
app_file: app.py
pinned: false
---
This space provides a gradio demo and an easy-to-run wrapper of the pre-trained model for structured sentiment analysis in Norwegian language, pre-trained on the [NoReC dataset](https://huggingface.co/datasets/norec).
This model is an implementation of the paper "Direct parsing to sentiment graphs" (Samuel _et al._, ACL 2022). The main repository that also contains the scripts for training the model, can be found on the project [github](https://github.com/jerbarnes/direct_parsing_to_sent_graph).
The current model uses the 'labeled-edge' graph encoding, and achieves the following results on the NoReC dataset:
| Unlabeled sentiment tuple F1 | Target F1 | Relative polarity precision |
|:----------------------------:|:----------:|:---------------------------:|
| 0.393 | 0.468 | 0.939 |
The model can be easily used for predicting sentiment tuples as follows:
```python
>>> import model_wrapper
>>> model = model_wrapper.PredictionModel()
>>> model.predict(['vi liker svart kaffe'])
[{'sent_id': '0',
'text': 'vi liker svart kaffe',
'opinions': [{'Source': [['vi'], ['0:2']],
'Target': [['svart', 'kaffe'], ['9:14', '15:20']],
'Polar_expression': [['liker'], ['3:8']],
'Polarity': 'Positive'}]}]
```
|