Scribd
Upload
15 views15 pages

Final Presentation

The Intelligent News Aggregator (INA) is an AI-based project aimed at addressing information overload through personalized content discovery and classification using machine learning and natural language processing. Key features include adaptive content classification, fake news detection, and an interactive user interface, with ongoing development in backend APIs and frontend enhancements. The project is being developed by a team from KCC Institute of Technology and Management and utilizes various technologies such as Python, Flask, and PostgreSQL.

Uploaded by

kriti pandey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
15 views15 pages

Final Presentation

The Intelligent News Aggregator (INA) is an AI-based project aimed at addressing information overload through personalized content discovery and classification using machine learning and natural language processing. Key features include adaptive content classification, fake news detection, and an interactive user interface, with ongoing development in backend APIs and frontend enhancements. The project is being developed by a team from KCC Institute of Technology and Management and utilizes various technologies such as Python, Flask, and PostgreSQL.

Uploaded by

kriti pandey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 15

Intelligent News

Aggregator
A full stack+ AI based
project Presented by :
Abhishek Kumar
Jha
Ayushi Tyagi
Fardin Alam
Harshit Kumar
Kriti Pandey
College:KCC Institute of Technology and Management
Course: B.Tech. (CSE)
Batch: 2021-25
OUR TEAM MEMBERS
Abhishe Ayushi
W O
Fardin
k Tyagi Alam
Kumar
ROLE: ROLE:
Jha
AI
ROLE : Modelling
Backend and
Database
Manageme
nt

Harshit Kriti
Kumar
ROLE:
Frontend and
T Pandey
ROLE:
Frontend and
Testing Documentation
ABOUT INTELLIGENT NEWS
AGGREGATOR
● The exponential growth of digital information has created
unprecedented challenges in content consumption and
knowledge acquisition . Information overload presents
significant cognitive barriers to effective content processing,
resulting in diminished user engagement and information
retention.
● This project proposes the development of an Intelligent News
Aggregator (INA) system that employs advanced machine
learning algorithms to facilitate personalized
• information discovery, classification, and presentation.
● By leveraging natural language processing , recommendation
systems ,
• user the INA aims to transform passive
and adaptive
consumption into
modeling, an optimized knowledge acquisition
information
process calibrated to individual user preferences, cognitive
patterns, and information needs.
K E Y FEATURES
• Adaptive content classification
• Personalized data recommendation
• Multi source data aggregation
• Intelligent content summarization
• Fake news detection and credibility scoring
• User profiling and interest analysis
• Interactive and intuitive visualization
• Scalable and efficient architecture
• Admin dashboard
Work In Progress
• Backend API Development
News retrieval APIs for frontend integration
User preference management
APIs Recommendation system optimization
• AI Podcast Feature
Text-to-speech conversion
Audio content organization
Delivery system integration
• Frontend Enhancements
User preference settings interface
News categorization view
Advanced filtering options
Email Service Issues Challenges
Project Architecture

Architecture Workflow :
1. Fetch RSS Feed
2. Summarize Description News
Aggregator Backend Report
3. Sentiment Analysis of Title
4. Named Entity Recognition (NER)
5. Save Final Processed News into
Database
6. Expose APIs via Flask
Technologies Used
• Python 3.11: Main backend language
• Flask: Web server to handle API routes
• PostgreSQL (Supabase): Database to store news data
• SQL: For creating database tables and queries
• Hugging Face Transformers: Pretrained AI models for summarization, sentiment,
NER
• Feedparser: Parse RSS feed (XML) into readable Python objects
• psycopg2-binary: PostgreSQL database driver for Python
• python-dotenv: Load environment variables securely
• Next.js : Main frontend framework
• Typescript: Main frontend language
Completed Work: Backend
• Core ML Models Implementation
BERT-based sentiment analysis model
User preference learning system
Text summarization model
• Data Pipeline
RSS feed integration for news collection
Database storage system for processed news
Processing pipeline through ML models
• Project Architecture
System design completed
Component interaction flow defined
Scalability considerations addressed
Database Table
id: UUID (Primary key)
(news):- title: TEXT
description: TEXT
summary: TEXT
sentiment_label: TEXT
sentiment_score: FLOAT
category: TEXT
published_at:
TIMESTAMP
source: TEXT
link: TEXT
image_url: TEXT
persons: TEXT
organizations: TEX
locations: TEXT
read_time: INT
popularity: IN
Current Backend Progress

RSS fetching: Done


AI summarization: Done
Sentiment analysis: Done
NER: Done
Database saving: Done
/api/fetch-news API:
Done
/api/news API: Planned
Completed Work: Frontend
• Authentication System
Password
recovery User
registration Login
functionality
• UI Components
Home page layout
(pre-authorization)
Personalized news
Current Frontend Progress

LOGIN PAGE SIGNUP PAGE


OUR HOMEPAGE

BEFORE LOGIN VEIW AFTER LOGIN VIEW


Future Steps

1. Complete API development


for
frontend integration
2.Resolve email service
production issues
3. Implement AI podcast
feature
4. Enhance user preference
system
5.Conduct comprehensive
system testing
6. Prepare for final deployment
Challenges Faced

Email services issues


• Working properly in development environment
• Not functioning in production environment
• Investigating SMTP configuration differences
• Exploring alternative email service providers
Other Technical Challenges
• Balancing model accuracy vs. performance
• Managing data volume for timely processing
• Maintaining database efficiency with growing datasets

You might also like