Uzaib52/aetheris-chat-dataset

Overview

This repository hosts the official benchmark and training dataset for Aetheris: AI-Powered Secure Offline Communication System.

The dataset is meticulously curated to train and evaluate on-device, lightweight text classifiers embedded within the Responsible Communication Framework (RCF). It empowers the Aetheris desktop application to proactively scan, filter, and classify message payloads (detecting telecom frauds, scams, and malicious content) completely offline before transmission over Bluetooth Low Energy (BLE) or RFCOMM sockets, ensuring absolute user privacy and data integrity.

---

Dataset Structure & Schema

The dataset comprises chat transmissions categorized into benign communication and structured high-risk threat vectors.

Data Fields

• message_id (string): Unique alphanumeric identifier assigned to each message packet.
• timestamp (string): ISO-8601 formatted timestamp of the message log generation.
• sender_id (string): Unique node or peer identifier within the local communication mesh.
• text (string): The raw textual payload of the message transmission.
• threat_category (string): Specific sub-classification of the risk vector (e.g., telecom_fraud, electricity_scam, or safe).
• risk_score (float): Normalized probability score ranging from 0.0 (benign) to 1.0 (critical threat).
• label (int): Binary target class (0 for Safe, 1 for High-Risk/Threat).

---

Quick Start: Loading in Python

You can easily ingest this dataset into your local machine learning pipeline or FastAPI backend using Hugging Face's datasets library:

from datasets import load_dataset

# Load the dataset from the Hugging Face Hub
dataset = load_dataset("Uzaib52/aetheris-chat-dataset")

# Convert to a Pandas DataFrame for exploratory analysis or preprocessing
df = dataset["train"].to_pandas()

# Display sample records
print(df.head())