albert-fa-base-v2-sentiment-digikala Open-source Model - Empowering Persian Language Understanding and Applications

Home

Albert Fa Base V2 Sentiment Digikala

Developed by m3hrdadfi

A lightweight BERT model for self-supervised language representation learning in Persian

Large Language Model

Transformers

OtherOpen Source License:Apache-2.0 #Persian Language Understanding #Lightweight BERT #Sentiment Analysis

Downloads 18

Release Time : 3/2/2022

Model Overview

ALBERT-Persian is the first attempt to implement the ALBERT model in the Persian language domain. Based on Google's ALBERT BASE 2.0, it is trained on multi-domain Persian corpora and suitable for Persian text understanding tasks.

Model Features

Lightweight Design

Based on the ALBERT architecture, parameter sharing mechanism makes the model more lightweight and efficient

Multi-domain Training

Trained on Persian corpora from various domains such as science, fiction, and news, ensuring broad coverage

Sentiment Analysis Optimization

Specifically optimized for sentiment classification tasks in Persian review texts

Model Capabilities

Persian Text Understanding

Sentiment Classification

Review Text Analysis

Use Cases

E-commerce Review Analysis

Digikala User Review Classification

Classify user reviews on the e-commerce platform Digikala into recommended/neutral/not recommended categories

F1 score reached 81.12%

Food Delivery Review Analysis

SnappFood User Review Analysis

Perform sentiment analysis on user reviews of the food delivery platform SnappFood

🚀 ALBERT Persian

A Lite BERT for Self-supervised Learning of Language Representations for the Persian Language

ALBERT-Persian is the first attempt on ALBERT for the Persian Language. The model was trained based on Google's ALBERT BASE Version 2.0 over various writing styles from numerous subjects (e.g., scientific, novels, news) with more than 3.9M documents, 73M sentences, and 1.3B words, similar to the approach used for ParsBERT.

Please follow the ALBERT-Persian repo for the latest information about previous and current models.

🚀 Quick Start

This project focuses on applying ALBERT to the Persian language for self - supervised learning of language representations. It also conducts sentiment analysis on Persian text using well - known datasets.

✨ Features

Persian Sentiment [Digikala, SnappFood, DeepSentiPers]

It aims to classify text, such as comments, based on their emotional bias. Three well - known datasets were tested for this task: Digikala user comments, SnappFood user comments, and DeepSentiPers in two binary - form and multi - form types.

Digikala

Digikala user comments are provided by Open Data Mining Program (ODMP). This dataset contains 62,321 user comments with three labels:

Label	#
no_idea	10394
not_recommended	15885
recommended	36042

Download You can download the dataset from here

📚 Documentation

Results

The following table summarizes the F1 score obtained as compared to other models and architectures.

Dataset	ALBERT - fa - base - v2	ParsBERT - v1	mBERT	DeepSentiPers
Digikala User Comments	81.12	81.74	80.74	-

BibTeX entry and citation info

Please cite in publications as the following:

@misc{ALBERTPersian,
  author = {Mehrdad Farahani},
  title = {ALBERT - Persian: A Lite BERT for Self - supervised Learning of Language Representations for the Persian Language},
  year = {2020},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/m3hrdadfi/albert-persian}},
}

@article{ParsBERT,
    title={ParsBERT: Transformer - based Model for Persian Language Understanding},
    author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
    journal={ArXiv},
    year={2020},
    volume={abs/2005.12515}
}

💬 Questions?

Post a Github issue on the ALBERT - Persian repo.

📄 License

This project is licensed under the Apache - 2.0 license.

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