politics - sentence - classifier open - source model - Free prediction of categories such as problems and solutions in French political sentences

Politics Sentence Classifier

Developed by mazancourt

This model is used to predict the nature category of French political statements, divided into problem, solution, and other.

Text Classification

Transformers

French#French Political Text Classification #Political Speech Analysis #Problem-Solution Identification

Downloads 14

Release Time : 3/2/2022

Model Overview

Trained on a mixed dataset of political tweets and speech excerpts, it accurately classifies the nature of French political statements (problem description, solution statement, or other types).

Model Features

Political Text Classification

Specifically designed for French political texts, capable of identifying problem descriptions, solution statements, and other types of content.

AutoNLP Training

Automatically trained and optimized using the AutoNLP platform, simplifying the model development process.

Environmental Consideration

The training process emitted only 1.06 grams of CO2, demonstrating environmental awareness.

Model Capabilities

French Text Classification

Political Statement Nature Identification

Problem Description Detection

Solution Statement Detection

Use Cases

Political Analysis

Political Speech Analysis

Analyze the content of political speeches to identify problems and solutions presented.

81% accuracy, effectively classifying the nature of political statements

Social Media Monitoring

Monitor politically related tweets and classify their content nature.

Media Studies

News Report Analysis

Analyze problem descriptions and solution statements in political news reports.

🚀 Prediction of sentence "nature" in a French political sentence

This model is designed to predict the nature of sentences in French political contexts. It classifies sentences into three categories, offering valuable insights into the content of political discourse.

✨ Features

Multi - class Classification: The model can categorize sentences into three distinct types: problem, solution, and other.
- problem: Sentences that describe an issue, such as il y a dans ce pays une fracture (J. Chirac).
- solution: Sentences presenting solutions, like J’ai supprimé les droits de succession parce que je crois au travail et parce que je crois à la famille. (N. Sarkozy).
- other: Sentences that don't fit into the above two categories, e.g., vive la République, vive la France.
Trained on Diverse Data: It was trained using AutoNLP with sentences from a combination of political tweets and speeches.

📦 Installation

No specific installation steps are provided in the original document, so this section is skipped.

📚 Documentation

Model Details

Property	Details
Problem type	Multi - class Classification
Model ID	23105051
CO2 Emissions (in grams)	1.06099358268878

Validation Metrics

Loss: 0.6050735712051392
Accuracy: 0.8097826086956522
Macro F1: 0.7713543865034599
Micro F1: 0.8097826086956522
Weighted F1: 0.8065488494385247
Macro Precision: 0.7861074705111403
Micro Precision: 0.8097826086956522
Weighted Precision: 0.806470454156932
Macro Recall: 0.7599656456873758
Micro Recall: 0.8097826086956522
Weighted Recall: 0.8097826086956522

💻 Usage Examples

Basic Usage

You can use cURL to access this model:

$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "Il y a dans ce pays une fracture"}' https://api-inference.huggingface.co/models/mazancourt/politics-sentence-classifier

Advanced Usage

Using the Python API:

from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained("mazancourt/autonlp-politics-sentence-classifier-23105051", use_auth_token=True)

tokenizer = AutoTokenizer.from_pretrained("mazancourt/politics-sentence-classifier", use_auth_token=True)

inputs = tokenizer("Il y a dans ce pays une fracture", return_tensors="pt")

outputs = model(**inputs)

# Category can be "problem", "solution" or "other"
category = outputs[0]["label"]
score = outputs[0]["score"]

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