AIbase
EN
Home
Korean Sentiment Analysis Kcelectra
K

Korean Sentiment Analysis Kcelectra

Developed by nlp04
A Korean sentiment analysis model fine-tuned based on KcELECTRA-base-v2022, which achieved good results on the evaluation set.
Text Classification
Transformers
Open Source License:MIT#Korean sentiment analysis#ELECTRA fine-tuning#High-precision text classification
Downloads 6,616
Release Time : 1/17/2023

Model Overview

This model is a Korean sentiment analysis model based on the Transformer architecture, mainly used to analyze the sentiment tendency of Korean texts.

Model Features

Korean sentiment analysis
Sentiment analysis ability optimized specifically for Korean texts
Fine-tuning based on KcELECTRA
Fine-tuned on the basis of KcELECTRA-base-v2022, inheriting the efficiency of the ELECTRA architecture
Multi-metric evaluation
Performs well on multiple metrics such as F1 score, AUPRC, and accuracy

Model Capabilities

Korean text classification
Sentiment tendency analysis
Natural language processing

Use Cases

Social media analysis
Sentiment analysis of Korean comments
Analyze the sentiment tendency of Korean comments on social media
The accuracy rate is approximately 70.7%
Customer feedback analysis
Classification of Korean customer feedback
Classify Korean customer feedback as positive/negative
The F1 score is approximately 70.7%
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase