license: mit language:

• en metrics:
• accuracy
• matthews_correlation widget:
• text: "Highway work zones create potential risks for both traffic and workers in addition to traffic congestion and delays that result in increased road user delay."
• text: "A circular economy is a way of achieving sustainable consumption and production, as well as nature positive outcomes."

sadickam/sdgBERT

sgdBERT (previously named "sdg-classification-bert"), is an NLP model for classifying text with respect to the United Nations sustainable development goals (SDG).

Source:https://www.un.org/development/desa/disabilities/about-us/sustainable-development-goals-sdgs-and-disability.html

Model Details

Model Description

This text classification model was developed by fine-tuning the bert-base-uncased pre-trained model. The training data for this fine-tuned model was sourced from the publicly available OSDG Community Dataset (OSDG-CD) Version 2023.10 at https://zenodo.org/records/8397907. This model was made as part of academic research at Deakin University. The goal was to make a transformer-based SDG text classification model that anyone could use. Only the first 16 UN SDGs supported. The primary model details are highlighted below:

• Model type: Text classification
• Language(s) (NLP): English
• License: mit
• Finetuned from model [optional]: bert-base-uncased

Model Sources

• Repository: https://huggingface.co/sadickam/sdg-classification-bert
• Demo: option 1 (copy/past text and csv): https://sadickam-sdg-text-classifier.hf.space/; option 2 (PDF documents): https://sadickam-document-sdg-app-cpu.hf.space

Direct Use

This is a fine-tuned model and therefore requires no further training.

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("sadickam/sdg-classification-bert")
model = AutoModelForSequenceClassification.from_pretrained("sadickam/sdg-classification-bert")

Training Data

The training data includes text from a wide range of industries and academic research fields. Hence, this fine-tuned model is not for a specific industry.

See training here: https://zenodo.org/records/8397907

Training Hyperparameters

• Num_epoch = 3
• Learning rate = 5e-5
• Batch size = 16

Evaluation

Metrics

• Accuracy = 0.90
• Matthews correlation = 0.89

Citation

Will be provided soon. Paper currently under review.

Model Card Contact

s.sadick@deakin.edu.au