AIbase
EN
Home
Irony Trained
I

Irony Trained

Developed by marcolatella
A text classification model fine-tuned on distilbert-base-uncased for identifying ironic content in tweets
Text Classification
Transformers
Open Source License:Apache-2.0#Twitter irony detection#Lightweight text classification#Social sentiment analysis
Downloads 19
Release Time : 3/2/2022

Model Overview

This model is a fine-tuned text classification model on the tweet_eval dataset, specifically designed to detect ironic expressions in tweets.

Model Features

Efficient and lightweight
Based on the DistilBERT architecture, it reduces model size while maintaining high performance
Irony detection
Specially optimized for detecting ironic expressions in social media text
Fine-tuning performance
Achieved an F1 score of 0.6946 on the tweet_eval dataset

Model Capabilities

Text classification
Irony detection
Social media text analysis

Use Cases

Social media analysis
Irony tweet detection
Automatically identify ironic content in social media
F1 score 0.6946
Enhanced sentiment analysis
Serves as a complementary component to sentiment analysis systems to improve recognition of ironic expressions
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