albert-sentiment Open-Source Traditional Chinese Sentiment Classification Model - Free and Precise Discrimination of Positive and Negative Sentiments

Albert Sentiment

Developed by clhuang

ALBERT pre-trained model fine-tuned for Traditional Chinese sentiment classification, supporting negative (0) and positive (1) sentiment classification.

Text Classification

Transformers

Other#Traditional Chinese Sentiment Analysis #ALBERT Fine-tuning #Short Text Classification

Downloads 217

Release Time : 6/5/2022

Model Overview

This model is used for sentiment classification of Traditional Chinese texts, determining whether the sentiment tendency is negative or positive.

Model Features

Traditional Chinese Support

Sentiment classification model specifically optimized for Traditional Chinese texts

Lightweight Model

Based on ALBERT architecture, more lightweight compared to traditional BERT models

Easy to Use

Provides clear API interfaces for easy integration into various applications

Model Capabilities

Traditional Chinese text sentiment classification

Negative/Positive sentiment judgment

Sentiment probability output

Use Cases

Social Media Analysis

Comment Sentiment Analysis

Analyze the sentiment tendency of user comments on social media

Quickly identify negative comments

Product Feedback Analysis

Product Review Classification

Perform sentiment classification on product reviews from e-commerce platforms

Help merchants understand customer satisfaction

🚀 Traditional Chinese Sentiment Classification: Negative(0), Positive(1)

This is an example model for a course, fine - tuned based on the ckiplab/albert pre - trained model. The training dataset only contains 80,000 samples.

🚀 Quick Start

This model is fine - tuned from the ckiplab/albert pre - trained model. It is used for sentiment classification in Traditional Chinese, distinguishing between negative (0) and positive (1) sentiments.

✨ Features

Based on ALBERT: Utilizes the ckiplab/albert pre - trained model.
Sentiment Classification: Classifies text into negative and positive sentiments.
Limited Training Data: Trained on a dataset of 80,000 samples, serving as an example model for courses.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("clhuang/albert-sentiment")
model = AutoModelForSequenceClassification.from_pretrained("clhuang/albert-sentiment")

## Pediction
target_names=['Negative','Positive']
max_length = 200 # Maximum number of characters. If it exceeds the number of characters during model training, the maximum number of characters of the model shall prevail.
def get_sentiment_proba(text):
    # prepare our text into tokenized sequence
    inputs = tokenizer(text, padding=True, truncation=True, max_length=max_length, return_tensors="pt")
    # perform inference to our model
    outputs = model(**inputs)
    # get output probabilities by doing softmax
    probs = outputs[0].softmax(1)

    response = {'Negative': round(float(probs[0, 0]), 2), 'Positive': round(float(probs[0, 1]), 2)}
    # executing argmax function to get the candidate label
    #return probs.argmax()
    return response

get_sentiment_proba('我喜歡這本書')
get_sentiment_proba('不喜歡這款產品')

📚 Documentation

The model is fine - tuned from the ckiplab/albert pre - trained model. The training dataset has 80,000 samples. It can classify Traditional Chinese text into negative and positive sentiments.

📄 License

This project is licensed under the AFL - 3.0 license.

Property	Details
Model Type	ALBERT - based sentiment classification model
Training Data	80,000 samples
Tags	albert, classification
Metrics	Accuracy

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご