System Requirements Document (SRD)

Project Name: infini-news

1. Introduction

The infini-news project aims to revolutionize the way users consume news by integrating advanced AI technologies to detect fake news and provide credibility insights. With misinformation being a global challenge, this system is designed to empower users with tools to verify authenticity, highlight misleading content, and evaluate source reliability. The project will cater to users in India (IN) and globally, ensuring locale-specific defaults such as Indian Standard Time (IST) and regional news prioritization where applicable.

This document outlines the system requirements for infini-news, including functional and non-functional specifications, user personas, design concepts, and technical architecture.

2. System Overview

infini-news is a web-based and Chrome Extension-supported platform that utilizes AI-driven Natural Language Processing (NLP) models to analyze news articles for authenticity. The system will provide:

• An authenticity score for pasted news content.
• Highlighting of misleading words or phrases.
• Evaluation of source credibility.
• Seamless integration with Chrome for real-time analysis of news directly on websites.

The backend will be powered by Node.js or Django, while the frontend will leverage React for a responsive and intuitive user interface. Advanced AI models like BERT and Logistic Regression will be employed for NLP tasks, and News APIs will provide real-time data for analysis.

3. Functional Requirements

Story Points:

• As a User, I should be able to paste news content into the system and receive an authenticity score.
• As a User, I should be able to see highlighted misleading words or phrases in the analyzed news.
• As a User, I should be able to view the credibility score of the news source.
• As a User, I should be able to install a Chrome Extension that analyzes news directly on websites.
• As an Admin, I should be able to manage AI models and update them for improved accuracy.
• As an Admin, I should be able to monitor system performance and user analytics.

4. User Personas

1. Everyday Reader

• Description: Individuals who consume news daily and want to verify its authenticity.
• Needs: Quick and reliable fake news detection, easy-to-use interface, and Chrome Extension support.

2. Journalist/Researcher

• Description: Professionals who require detailed analysis of news content for reporting or research purposes.
• Needs: In-depth credibility scoring, highlighted misleading content, and access to source reliability metrics.

3. Admin

• Description: System administrators responsible for maintaining and improving the platform.
• Needs: Tools to manage AI models, monitor system performance, and ensure data security.

5. Visuals Colors and Theme

Color Palette:

• Background: #F5F5F5 (Soft Gray)
• Surface: #FFFFFF (Pure White)
• Text: #333333 (Dark Charcoal)
• Accent: #FF5722 (Vivid Orange)
• Muted Tones: #B0BEC5 (Cool Blue-Gray)

Theme:

The theme will be modern and professional, with a focus on clarity and trustworthiness. The accent color will be used sparingly to highlight key features like authenticity scores and credibility metrics.

6. Signature Design Concept

Concept: Interactive News Analyzer Dashboard

The homepage will feature a dynamic dashboard that visually represents the analysis of news articles. Users will see:

• Authenticity Score Gauge: A circular gauge that animates from 0 to the calculated score, with color transitions (green for authentic, yellow for questionable, red for fake).
• Word Cloud Visualization: Misleading words will appear in a live word cloud, with larger fonts for higher impact words. Clicking on a word will display its context in the article.
• Source Credibility Bar: A horizontal bar that fills dynamically based on the credibility score, with micro-interactions like tooltips explaining the score.
• Chrome Extension Preview: A live demo section showing how the extension works on popular news websites.

Animations will include smooth transitions, hover effects, and subtle color shifts to enhance user engagement. The design will evoke a sense of transparency and empowerment, making users feel in control of their news consumption.

7. Non-Functional Requirements

• Performance: The system must analyze and return results within 3 seconds for pasted news content.
• Scalability: Support up to 1 million users concurrently.
• Security: Implement robust encryption for user data and AI model integrity.
• Availability: Ensure 99.9% uptime with failover mechanisms.
• Localization: Provide support for Indian regional languages and prioritize Indian news sources.

8. Tech Stack

Frontend

• React for Web
• React Native for mobile app

Backend

• Node.js or Django
• FastAPI

Database

• MySQL or MariaDB for structured data
• MongoDB for unstructured data

AI Models

• BERT for NLP tasks
• Logistic Regression for classification

AI Tools

• Litellm for LLM Routing
• Langchain

Local Orchestration

• Docker
• docker-compose

Server-side Orchestration

• Kubernetes

9. Assumptions and Constraints

Assumptions

• Users will primarily access the system via web browsers and Chrome Extensions.
• News APIs will provide reliable and up-to-date data for analysis.
• The AI models will be trained on diverse datasets to ensure accuracy across multiple languages and regions.

Constraints

• Limited initial support for non-English languages outside India.
• Dependency on third-party News APIs for real-time data.
• Chrome Extension functionality may vary based on browser updates.

10. Glossary

• Authenticity Score: A numerical representation of the likelihood that a news article is genuine.
• Misleading Words: Words or phrases identified by the AI as potentially deceptive or manipulative.
• Source Credibility: A metric evaluating the trustworthiness of the news source based on historical data and reliability.
• NLP (Natural Language Processing): A branch of AI focused on enabling computers to understand and process human language.
• BERT: Bidirectional Encoder Representations from Transformers, an advanced NLP model.
• Chrome Extension: A browser add-on that integrates the system's functionality directly into the user's browsing experience.