Multilingual DeBERTa V3 Base for Sentiment Assessment

This is a fine-tuned version of the multilingual DeBERTa model (mdeberta) for assessing text sentiment across languages.

Model Details

• Architecture: mdeberta-v3-base-sentiment
• Task: Classification (Sentiment Analysis)
• Training Data: agentlans/tatoeba-english-translations containing 48 900 labeled English translations
• Input: Text in any of the supported languages by DeBERTa
• Output: Sentiment score for text (positive, negative, neutral)
  • positive scores indicate a positive sentiment
  • zero score indicate neutral sentiment
  • negative scores indicate a negative sentiment

Performance

RMSE accuracy on 20% held-out validation set: 0.4177

Training Data

The model was trained on agentlans/tatoeba-english-translations.

Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model_name="agentlans/mdeberta-v3-base-sentiment"

# Put model on GPU or else CPU
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = model.to(device)

def sentiment(text):
    """Processes the text using the model and returns its logits.
    In this case, it's interpreted as the sentiment score for that text."""
    inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
    with torch.no_grad():
        logits = model(**inputs).logits.squeeze().cpu()
    return logits.tolist()

# Note: Recommend to preprocess text to remove special characters, e-mails, and hash tags
sentiment("Your text here.")

Results

In this study, 10 English text samples of varying sentiment were generated and translated into Arabic, Chinese, French, Russian, and Spanish using Google Translate. This resulted in a total of 50 translated samples, which were subsequently analyzed by a trained classifier to predict their sentiment scores.

The scatterplot below illustrates the predicted sentiment scores grouped by each text sample. Notably, the prediction scores exhibit low variability across different languages for the same text, indicating a consistent assessment of translation sentiment regardless of the target language.

This analysis highlights the effectiveness of using machine learning classifiers in evaluating textual sentiment across multiple languages.

Limitations

• Performance may vary for texts significantly different from the training data
• Output is based on statistical patterns and may not always align with human judgment
• Sentiment is assessed purely on textual features, not considering factors like subject familiarity or cultural context

Ethical Considerations

• Should not be used as the sole determinant of text suitability for specific audiences
• Results may reflect biases present in the training data sources
• Care should be taken when using these models in educational or publishing contexts