INSTITUTE OF ENGINEERING AND MANAGEMENT
KOLKATA
YEAR : 2022-2026
Name : TANMOY SAHA
SEMESTER : 5TH
ROLL NO. : 56
DEPARTMENT : INFORMATION TECHNOLOGY(IT)
ENROLMENT NO. : 12022002004070
SUBJECT : AIML LAB
PROJECT : 360-degree feedback software for the
Government of India related News Stories in
Regional Media using Artificial Intelligence /
Machine Learning
1|Page
INDEX
S.NO. TOPIC PAGE NO.
1 ACKNOWLEDGEMENT 3
2 PROJECT DEFINATION 4
3 METHODOLOGY 5
4 RESULTS 6-7
5 MY CONTRIBUTION TO 8
PROJECT
6 DISCUSSION 9
7 REFERENCE 10
2|Page
Acknowledgement
I wish to express my heartfelt gratitude to the all the people
who have played a crucial role in the research for this
project, without their active cooperation the preparation of
this project could not have been completed within the
specified time limit. I am thankful to my project Guide(s)
Prof. Sanchita Chakraborty who supported me throughout
this project with utmost cooperation and patience and for
helping me in doing this Project. I am also thankful to our
respected Head of the Department, Prof. Dr. Moutushi
Singh, for motivating me to complete this project with
complete focus and attention. I am thankful to my
department and all my teachers for the help and guidance
provided for this work.
3|Page
DEFINATION
360-degree feedback software for the Government of India, utilizing
Artificial Intelligence (AI) and Machine Learning (ML), is a system designed
to collect, analyze, and interpret feedback or insights from various
stakeholders. The stakeholders include government officials, employees,
and citizens, and the data is gathered from regional media sources like
newspapers, TV channels, online portals, and social media platforms.
The software integrates AI/ML technologies to process multilingual and
regional content, detect sentiments, identify trends, and provide actionable
recommendations. It aims to enhance transparency, accountability, and
efficiency in government operations and decision-making by gathering
holistic, real-time perspectives on policies, initiatives, and public
sentiments.
KEY FEATURES :-
Multilingual Support: Ability to analyze feedback in multiple regional
languages.
Sentiment Analysis: Automated understanding of positive, neutral, or
negative public sentiment.
Media Monitoring: Continuous scanning of regional news and social
media platforms for relevant information.
Stakeholder Engagement: Collects inputs from various levels (public,
employees, officials).
Insights and Reporting: Provides dashboards and reports for policy-
makers to make data-driven decisions.
Trend Analysis: Identifies emerging issues or popular narratives in
regional media.
4|Page
METHODOLOGY
Data Collection :- Gather data from regional newspapers,
television channels, radio transcripts, online news portals, and
social media platforms. Incorporate tools to scrape and process
content in multiple Indian regional languages. Use APIs or
scraping tools to fetch real-time data from digital platforms.
Data Preprocessing :- Clean raw data by removing irrelevant
content (e.g., advertisements, noise). Use Natural Language
Processing (NLP) models for translation, if needed, while retaining
the original sentiment and context. Identify and remove duplicate
content from multiple sources. Organize data into predefined
categories like health, education, economy, etc.
AI/ML Models for Analysis :- Use supervised ML models (e.g.,
SVM, Random Forest) or deep learning models (e.g., BERT, GPT)
to classify sentiments (positive, neutral, negative). Fine-tune
models on regional language datasets for accurate analysis.
Employ algorithms like Latent Dirichlet Allocation (LDA) to identify
trending topics.
Feedback Integration :- Collect feedback from government officials
and employees alongside public feedback from media data.
Compare AI-detected insights with human feedback for accuracy.
Reporting and Visualization :- Develop user-friendly, interactive
dashboards to present real-time insights, trends, and sentiment
scores. Set up alerts for significant changes in public sentiment or
coverage of government policies. Highlight critical issues requiring
immediate attention.
Continuous Learning and Model Improvement :- Incorporate user
feedback to improve AI/ML models over time. Regularly update
datasets to include new languages, slang, or terminology. Ensure
models are robust to biases and adapt to evolving media trends.
Conduct audits to verify the accuracy and relevance of insights.
5|Page
RESULTS
Comprehensive Media Monitoring :- Continuous tracking of news
stories, opinions, and trends in regional media. Provides a
centralized repository of all regional news relevant to government
initiatives. Captures feedback from a wide range of stakeholders
across different languages and regions.
Enhanced Sentiment Analysis :- Accurate measurement of public
sentiment towards government policies, schemes, and actions.
Differentiation of sentiments by demographics, geography, or
specific schemes. Quick identification of negative sentiment trends
or dissatisfaction, allowing timely interventions.
Data-Driven Policy Recommendations:- Actionable insights into
the effectiveness of government policies and their reception. Data-
backed recommendations for modifications or communication
strategies. Evaluation of policy impact in terms of public
awareness, acceptance, and benefits.
Transparency and Accountability:- Ensures government
accountability by addressing media-highlighted concerns and
sentiments. Strengthens trust between the government and the
public. Clear documentation of how feedback influences policy
adjustments.
Identification of Regional Trends and Issues:- Pinpoints regional
challenges and concerns highlighted in local media. Offers
opportunities for targeted interventions in specific states or
districts.Identification of new or underreported issues requiring
immediate attention.
Enhanced Multilingual and Regional Capabilities:- Adapts insights
to account for linguistic nuances and cultural differences. Tailored
communication strategies for diverse regions.
Efficiency Gains:- Automation of large-scale media analysis
reduces manual effort and cost. Real-time updates minimize delays
in recognizing critical issues.
6|Page
Efficiency Gains:- Automation of large-scale media analysis
reduces manual effort and cost. Real-time updates minimize
delays in recognizing critical issues.
Improved Crisis Management:- Early alerts for crises based on
anomalies or spikes in negative sentiment.Effective management
of public relations and mitigation of negative publicity.
Public Perception Tracking Over Time:- Tracks changes in public
sentiment across the lifecycle of a policy or initiative.Assists in
understanding the long-term effects of government actions.
Highlights positive outcomes and success stories from regional
media coverage.
Better Stakeholder Collaboration:- Facilitates communication
among government departments by sharing insights.Encourages
collective action to address public feedback. Promotes
participatory governance by including citizen feedback in
decision-making.
7|Page
My contribution To Project
• Collaborate with the team to define specific goals and key
performance indicators (KPIs).
• Suggest best practices in data collection and interpretation.
•Develop or train machine learning models for multilingual
sentiment analysis.
• Create tools for keyword extraction or context understanding.
• Design or test front-end interfaces.
• Propose strategies to audit and validate model fairness.
• Work on secure data storage and processing methods.
• Participate in pilot testing and suggest optimizations.
• Design scalability strategies, considering technical and
logistical constraints. • Organize in training sessions.
• Prepare concise, clear documentation for users.
8|Page
DISCUSSION
Importance and Relevance :- Regional media often captures
grassroots-level issues and sentiments that may not be visible in
mainstream media. This software ensures that the voices of
citizens from remote and rural areas are considered in
policymaking.
Timely Decision-Making:- Real-time insights from AI/ML models
enable policymakers to react swiftly to emerging issues,
minimizing the impact of public dissatisfaction or crises. By
systematically monitoring media feedback, the government can
demonstrate responsiveness to public concerns, boosting trust
and credibility.
Use Cases:- Evaluate how specific policies are perceived across
different regions and demographics. Detect and respond to
controversies, protests, or dissatisfaction early. Refine
communication strategies by understanding public sentiment and
preferred messaging tones in different regions.
Technical and Implementation Challenges:- India’s linguistic
diversity makes it challenging to accurately process text and
speech in multiple regional languages. Advanced Natural
Language Processing (NLP) models trained on regional datasets
are necessary.
Opportunities and Benefits:- Enables the government to
incorporate feedback from diverse regions and communities,
ensuring no group is left unheard. Helps policymakers transition
from intuition-based to evidence-based decision-making.
AI/ML Role:- Processes and analyzes text in multiple languages to
extract insights from media content. Identifies patterns and
predicts emerging issues or public trends. Reduces manual effort
in media monitoring, allowing human resources to focus on
interpretation and action.
9|Page
REFERENCE
Patel, D., & Kumar, R. (2021). "Challenges in NLP for Indian
Languages: Sentiment Analysis and Beyond."
Singh, A., & Sharma, P. (2020). "Role of Artificial Intelligence in
Enhancing Public Administration."
Gupta, S., et al. (2022). "Machine Learning Models for Sentiment
Analysis of Indian News Articles."
10 | P a g e
