code for our NAACL 2019 paper "BERT Post-Training for Review Reading Comprehension and Aspect-based Sentiment Analysis".
Preprocessing ABSA xmls organized into a separate rep.
Want to have post-trained models for other domains in reviews ? checkout a cross-domain review BERT or download from HERE.
A conversational dataset of RRC can be found here.
If you only care about ASC, a more formal code base can be found in a similar rep focusing on ASC.
**feedbacks are welcomed for missing instructions **
We focus on 3 review-based tasks: review reading comprehension (RRC), aspect extraction (AE) and aspect sentiment classification (ASC).
RRC: given a question ("how is the retina display ?") and a review ("The retina display is great.") find an answer span ("great") from that review;
AE: given a review sentence ("The retina display is great."), find aspects("retina display");
ASC: given an aspect ("retina display") and a review sentence ("The retina display is great."), detect the polarity of that aspect (positive).
The code is tested on Ubuntu 16.04 with Python 3.6.8(Anaconda), PyTorch 1.0.1 and pytorch-pretrained-bert 0.4. (sorry for the extreme old version of this library, we will release a newer version in the first half of 2020.)
We suggest make an anaconda environment for all packages and uncomment environment setup in script/run_rrc.sh script/run_absa.sh script/pt.sh.
The post-training code additionally use apex 0.1 to speed up training on FP16, which is compiled with PyTorch 1.0.1(py3.6_cuda10.0.130_cudnn7.4.2_2) and CUDA 10.0.130 on RTX 2080 Ti. It is possible to avoid use GPUs that do not support apex (e.g., 1080 Ti), but need to adjust the max sequence length and number of gradient accumulation but (although the result can be better).
Fine-tuning code is tested without using apex 0.1 to ensure stability.
Our evaluation wrapper code is written in ipython notebook eval/eval.ipynb.
But you are free to call the evaluation code of each task separately.
AE eval/evaluate_ae.py additionally needs Java JRE/JDK to be installed.
step1: make 2 folders for post-training and fine-tuning.
mkdir -p pt_model ; mkdir -p run
step2: place post-trained BERTs into pt_model/. Our post-trained Laptop weights can be download here and restaurant here. You are free to download other BERT weights into this folder(e.g., bert-base, BERT-DK (laptop, restaurant) in our paper). Make sure to add an entry into src/modelconfig.py.
step3: make 3 folders for 3 tasks:
place fine-tuning data to each respective folder: rrc/, ae/, asc/. A pre-processed data in json format is in data/json_data.tar.gz, or can be downloaded here.
step4: fire a fine-tuning from a BERT weight, e.g.
cd script
bash run_rrc.sh rrc laptop_pt laptop pt_rrc 10 0
Here rrc is the task to run, laptop_pt is the post-trained weights for laptop, laptop is the domain, pt_rrc is the fine-tuned folder in run/, 10 means run 10 times and 0 means use gpu-0.
similarly,
bash run_rrc.sh rrc rest_pt rest pt_rrc 10 0
bash run_absa.sh ae laptop_pt laptop pt_ae 10 0
bash run_absa.sh ae rest_pt rest pt_ae 10 0
bash run_absa.sh asc laptop_pt laptop pt_asc 10 0
bash run_absa.sh asc rest_pt rest pt_asc 10 0
step5: evaluation
RRC: download SQuAD 1.1 evaluation script (e.g. ) to eval/.
AE: place official evaluation .jar files as eval/A.jar and eval/eval.jar.
place testing xml files as (the step 4 of this has a similar setup)
ae/official_data/Laptops_Test_Gold.xml
ae/official_data/Laptops_Test_Data_PhaseA.xml
ae/official_data/EN_REST_SB1_TEST.xml.gold
ae/official_data/EN_REST_SB1_TEST.xml.A
ASC: built-in as part of eval/eval.ipynb
open result.ipynb and run as you wish
Assume you are on the root folder of this repository.
step1: domain post-training data:
Download pre-processed data from here and place it under the root folder.
Alternatively you can create your own data. For example for laptop, mkdir -p domain_corpus ; mkdir -p domain_corpus/laptop ; mkdir -p domain_corpus/raw. Place newline separated reviews laptop.txt as domain_corpus/raw/laptop.txt
The script in step3 will detect if your .npz is unavailable under domain_corpus/laptop then it will start to preprocess domain_corpus/raw/laptop.txt first.
step2: MRC post-training data:
Place and rename SQuAD 1.1 (for consistency with other review-based tasks) as squad/train.json.
step3: preprocessing and post-training
bash pt.sh laptop 5 70000 0
bash pt.sh rest 1 140000 0
If you find this work useful, please cite as following.
@inproceedings{xu_bert2019,
title = "BERT Post-Training for Review Reading Comprehension and Aspect-based Sentiment Analysis",
author = "Xu, Hu and Liu, Bing and Shu, Lei and Yu, Philip S.",
booktitle = "Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics",
month = "jun",
year = "2019",
}