English|简体中文
Senta is a python library for many sentiment analysis tasks. It contains support for running multiple tasks such as sentence-level sentiment classification, aspect-level sentiment classification and opinion role labeling. The bulk of the code in this repository is used to implement SKEP: Sentiment Knowledge Enhanced Pre-training for Sentiment Analysis. In the paper, we demonstrate how to integrate sentiment knowledge into pre-trained models to learn a unified sentiment representation for multiple sentiment analysis tasks.
You can directly use the Python package to predict sentiment analysis tasks by loading a pre-trained SKEP model.
-
Sentasupports Python 3.6 or later. This repository requires PaddlePaddle 1.6.3, please see here for installaton instruction. -
Install
Sentapython -m pip install Senta
or
git clone https://github.com/baidu/Senta.git cd Senta python -m pip install .
```python
from senta import Senta
my_senta = Senta()
# get pre-trained model, we provide three pre-trained models, all of which are based on the SKEP
print(my_senta.get_support_model()) # ["ernie_1.0_skep_large_ch", "ernie_2.0_skep_large_en", "roberta_skep_large_en"]
# ernie_1.0_skep_large_ch, skep Chinese pre-trained model based on ERNIE 1.0 large.
# ernie_2.0_skep_large_en, skep English pre-trained model based on ERNIE 2.0 large.
# roberta_skep_large_en, skep English pre-trained model based on RoBERTa large, which is used in our paper.
# get supported task
print(my_senta.get_support_task()) # ["sentiment_classify", "aspect_sentiment_classify", "extraction"]
use_cuda = True # set True or False
# predict different tasks
my_senta.init_model(model_class="roberta_skep_large_en", task="sentiment_classify", use_cuda=use_cuda)
texts = ["a sometimes tedious film ."]
result = my_senta.predict(texts)
print(result)
my_senta.init_model(model_class="roberta_skep_large_en", task="aspect_sentiment_classify", use_cuda=use_cuda)
texts = ["I love the operating system and the preloaded software."]
aspects = ["operating system"]
result = my_senta.predict(texts, aspects)
print(result)
my_senta.init_model(model_class="roberta_skep_large_en", task="extraction", use_cuda=use_cuda)
texts = ["The JCC would be very pleased to welcome your organization as a corporate sponsor ."]
result = my_senta.predict(texts)
print(result)
```
You can use the source code to run pre-training and fine-tuning tasks. The config folder has different files to help you reproduce the results of our paper.
```shell
# download code
git clone https://github.com/baidu/Senta.git
# download a pre-trained skep model
cd ./Senta/model_files
sh download_roberta_skep_large_en.sh # download roberta_skep_large_en model. For other pre-trained skep models, you can find them in this dir.
cd -
# download task dataset
cd ./Senta/data/
sh download_en_data.sh # download English dataset used in our paper. For Chinese dataset, you can find its download script in this dir.
cd -
```
-
Sentasupports Python 3.6 or later. This repository requires PaddlePaddle 1.6.3, please see here for installaton instruction. -
Install python dependencies
python -m pip install -r requirements.txt
-
Set up environment variables such as Python, CUDA, cuDNN, PaddlePaddle in
env.shfile. Details about environment variables related to PaddlePaddle can be found at the PaddlePaddle Documentation.
-
Training
sh ./script/run_pretrain_roberta_skep_large_en.sh # pre-trained model roberta_skep_large_en, which is used in our paper -
Fine-tuning and predict
sh ./script/run_train.sh ./config/roberta_skep_large_en.SST-2.cls.json # fine-tuning on SST-2 sh ./script/run_infer.sh ./config/roberta_skep_large_en.SST-2.infer.json # predict sh ./script/run_train.sh ./config/roberta_skep_large_en.absa_laptops.cls.json # fine-tuning on ABSA(laptops) sh ./script/run_infer.sh ./config/roberta_skep_large_en.absa_laptops.infer.json # predict sh ./script/run_train.sh ./config/roberta_skep_large_en.MPQA.orl.json # fine-tuning on MPQA 2.0 sh ./script/run_infer.sh ./config/roberta_skep_large_en.MPQA.infer.json # predict
-
An old version of
Sentacan be found at here, which includes BoW, CNN and BiLSTM models for Chinese sentence-level sentiment classification.
If you extend or use this work, please cite the paper where it was introduced:
@inproceedings{tian-etal-2020-skep,
title = "{SKEP}: Sentiment Knowledge Enhanced Pre-training for Sentiment Analysis",
author = "Tian, Hao and
Gao, Can and
Xiao, Xinyan and
Liu, Hao and
He, Bolei and
Wu, Hua and
Wang, Haifeng and
wu, feng",
booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
month = jul,
year = "2020",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://www.aclweb.org/anthology/2020.acl-main.374",
pages = "4067--4076",
abstract = "Recently, sentiment analysis has seen remarkable advance with the help of pre-training approaches. However, sentiment knowledge, such as sentiment words and aspect-sentiment pairs, is ignored in the process of pre-training, despite the fact that they are widely used in traditional sentiment analysis approaches. In this paper, we introduce Sentiment Knowledge Enhanced Pre-training (SKEP) in order to learn a unified sentiment representation for multiple sentiment analysis tasks. With the help of automatically-mined knowledge, SKEP conducts sentiment masking and constructs three sentiment knowledge prediction objectives, so as to embed sentiment information at the word, polarity and aspect level into pre-trained sentiment representation. In particular, the prediction of aspect-sentiment pairs is converted into multi-label classification, aiming to capture the dependency between words in a pair. Experiments on three kinds of sentiment tasks show that SKEP significantly outperforms strong pre-training baseline, and achieves new state-of-the-art results on most of the test datasets. We release our code at https://github.com/baidu/Senta.",
}