InstructABSA: Instruction learning for Aspect Based Sentiment Classification

(NOTE: This repository is a work in progress it will be completed by Sunday - February 26 2023)

This repository is the official implementation of the paper. As part of our approach we show the efficacy of instruction tuned language models. This approach surpasses previous SOTA on downstream ABSA subtasks by significant margin.

Dataset Requirements

This section describes the format of the data required for the training and evaluation of the datasets using our approach. For all subtasks, the field names should match exactly as shown and have the same datatypes. The fields to be present in the raw dataset are as follows:

-- raw_text: This is the reviews section (str)

-- aspectTerms: This is the set of aspect terms and their polarities to be present as a list of dictionaries. Each dictionary will have atleast two keys with the one of the key term and the value which is an aspect in the corresponding sentence. The second key will be polarity and its value is the polarity for corresponding aspect. ([{'term':'aspect1', 'polarity':'polarity1'}, ...])

Warning When creating the dataset in this fashion and saving it, .xlsx/.csv format will convert the aspectTerms column into string/text format. But the package will handle that when loading the dataset file.

An example dataset is shown below:

raw_text	aspectTerms
The cab ride was amazing but the service was pricey	[{'term':'cab ride', 'polarity':'positive'}, {'term':'service', 'polarity':'negative'}]
I ordered the Barbeque Pizza	[]

Model Checkpoints (Coming Soon)

Task	Model Name	Remarks
ATE	-	-
ATSC	-	-

Aspect Term Extraction

The ATE models can be trained from scratch or alternatively can be used to run inference on your datasets directly. There are two ways this can be done. The first one is through the CLI commands which is shown below:

To evaluate the ATE subtask on a single input using CLI run the following:

python run_model.py -mode cli -task ate \
-model_checkpoint Models/ATE/allenai/tk-instruct-base-def-pos-combined100_instruct_pos_neg_neut/checkpoints \
-test_input 'The cab ride was amazing but the service was pricey'

To run the same using the InstructABSA module, the steps are described below:

from InstructABSA.utils import T5Generator

Aspect Term Sentiment Classification

The ATE models can be trained from scratch or alternatively can be used to run inference on your datasets directly. There are two ways this can be done. The first one is through the CLI commands which is shown below:

To evaluate the ATSC subtask on a single input using CLI run the following:

python run_model.py -mode cli -task atsc \
-model_checkpoint Models/ATSC/allenai/tk-instruct-base-def-pos-combined100_instruct_pos_neg_neut/checkpoints \
-test_input 'The cab ride was amazing but the service was pricey|cab ride'

Note the | delimiter that is used to pass the aspect term for which the polarity is to be extracted.

To run the same using the InstructABSA module, the steps are described below:

from InstructABSA.utils import T5Classifier

Joint Tasks

The ATE models can be trained from scratch or alternatively can be used to run inference on your datasets directly. There are two ways this can be done. The first one is through the CLI commands which is shown below:

To evaluate the Joint Task on a single input using CLI run the following:

python run_model.py -mode cli -task joint \
-model_checkpoint Models/JointTask/allenai/tk-instruct-base-def-pos-combined100_instruct_pos_neg_neut/checkpoints \
-test_input 'The cab ride was amazing but the service was pricey'

To run the same using the InstructABSA module, the steps are described below:

from InstructABSA.utils import T5Generator