Social network security unit 3

UNIT – I
Introduction to the Semantic Web
The Semantic Web is an extension of the current World Wide Web (WWW) that aims to make data machine-
readable and meaningfully linked. It enables computers to understand, interpret, and process web content in a way
that enhances data integration, automation, and artificial intelligence (AI) applications.

Key Idea:
The current web is mainly designed for humans (text, images, videos).
The Semantic Web makes content structured and linked, allowing machines to process and understand it
intelligently.
Applications of the Semantic Web
A. Search Engines (Google Knowledge Graph)
Google Knowledge Graph links information, improving search accuracy.
Example: Searching for "Barack Obama" gives a knowledge box with structured data about him.
B. Personal Assistants (Siri, Alexa, Google Assistant)
Understands user intent by analyzing structured semantic data.
C. Healthcare and Scientific Research
Medical ontologies link diseases, drugs, and symptoms.
Example: IBM Watson uses Semantic Web to recommend treatments.
D. E-commerce and Recommendation Systems
Amazon, Netflix use linked data to provide personalized recommendations.
Advantages of the Semantic Web
 Better Search Accuracy – Context-aware search results.
 Data Integration – Connects diverse datasets across different sources.
 Automation & AI – Enables intelligent applications.
 Interoperability – Standardized formats allow seamless data exchange.

Challenges of the Semantic Web

 Complexity – Requires structured data and ontology design.
 Scalability – Processing large RDF datasets is computationally expensive.
 Adoption Issues – Many websites still rely on unstructured HTML content.

Limitations of the Current Web

The current Web (Web 2.0) has revolutionized communication, information sharing, and commerce, but it also
has several limitations related to data management, security, privacy, and intelligence. Below are some key
challenges:

1. Lack of Machine Understanding (Limited Intelligence)

The Web is designed for humans, not machines.
Search engines rely on keywords, not understanding context.
Example: A search for “apple” may return results for both Apple Inc. and the fruit without distinguishing meaning.
How the Semantic Web Improves This
The Semantic Web (Web 3.0) introduces metadata, ontologies, and linked data to give meaning to web content.
2. Poor Data Integration & Interoperability
Different websites use different data formats (JSON, XML, HTML, CSV).
Example: A hospital system and a fitness tracker may store health data in incompatible formats.
Why This Is a Problem
No standardized way to share or integrate data across platforms.
Data silos prevent seamless information exchange.
Potential Solution
The Semantic Web enables structured data formats (RDF, OWL, Linked Data) for better integration.
3. Privacy & Security Concerns
Web 2.0 relies on centralized platforms (Google, Facebook, Amazon), which collect and control vast amounts of
personal data.
Issues:
Mass surveillance (Governments & Corporations track user activity).
Data breaches (Facebook, Equifax leaks).
Lack of user control (Users don’t own their data).
Possible Solution: Decentralized Web (Web 3.0)
Blockchain-based Web enables self-sovereign identity and user-controlled data.
4. Fake News, Misinformation & Lack of Content Verification
Anyone can publish information without fact-checking.
Examples:
Fake news on social media influences elections.
AI-generated deepfakes mislead people.
Solution
AI-powered fact-checking systems integrated into search engines.
Blockchain-based verification for digital content authenticity.
5. Centralization & Monopoly Power
A few companies control most of the web:
Google (Search & Ads)
Facebook (Social Media)
Amazon (E-commerce & Cloud)
These companies influence what users see (filter bubbles, censorship).
Alternative: Decentralized Web (Web3)
Peer-to-peer (P2P) networks like IPFS (InterPlanetary File System) can reduce reliance on centralized servers.
6. Slow & Inefficient Web Performance
Loading speeds depend on central servers, leading to:
High latency in remote regions.
Server crashes affecting services.
Example: If AWS (Amazon Web Services) goes down, major websites crash.
Solution: Distributed Web Infrastructure
Edge computing and content delivery networks (CDNs) optimize web performance.
7. Lack of Personalization & Adaptive Interfaces
The Web does not fully adapt to individual preferences and behaviors.
Example: A visually impaired person may struggle to access web content due to poor accessibility features.
Solution: AI-Driven Adaptive Web
AI and Natural Language Processing (NLP) can improve personalized content delivery.
Conclusion
The current Web has several limitations, but the evolution towards Web 3.0, the Semantic Web, and decentralized
technologies offers potential solutions to create a more intelligent, secure, and user-controlled web.
Development of the Semantic Web
The Semantic Web is an evolution of the World Wide Web (WWW) aimed at making web data machine-readable,
structured, and interconnected. It enables AI-driven automation, better search results, and improved data integration
across different platforms.

Evolution of the Web

A. Web 1.0 (Static Web) – 1990s

The first generation of the web.
Read-only websites with static content (e.g., early Yahoo, Britannica).
No interactivity, no user-generated content.
B. Web 2.0 (Social & Interactive Web) – 2000s-Present
User-generated content (social media, blogs, wikis).
Platforms like Facebook, Twitter, YouTube, and Wikipedia.
Challenges:
Data silos (different platforms store data separately).
No machine understanding of web content.
Privacy and security concerns.
C. Web 3.0 (Semantic & Intelligent Web) – Emerging
Goal: Enable machines to understand, interpret, and connect web data.
Uses AI, linked data, ontologies, and metadata to structure information.
Applications in search engines, chatbots, digital assistants (Siri, Alexa).

Social Network Analysis (SNA)

1. Introduction to Social Network Analysis (SNA)
Social Network Analysis (SNA) is a method for examining relationships and interactions among individuals, groups,
organizations, or systems. It applies graph theory and mathematical models to analyze structures, patterns, and
dynamics in social networks.

Key Concepts in SNA

Nodes (Vertices) → Represent individuals, organizations, or entities.
Edges (Links) → Represent relationships or interactions between nodes.
Graph → A collection of nodes and edges forming a network.
SNA helps in understanding influence, information flow, community structures, and network dynamics in various
fields such as social media, business, healthcare, and cybersecurity.

2. Applications of Social Network Analysis

A. Social Media & Online Communities
 Identifying influencers (e.g., top Twitter users with high engagement).
 Detecting fake news propagation and bot networks.
 Understanding viral trends and content spread.

B. Business & Marketing

 Targeted advertising based on network connections.
 Customer segmentation for better marketing strategies.
 Recommendation systems (Netflix, Amazon product recommendations).

C. Political & Social Movements

 Analysis of activist networks (e.g., Arab Spring, #MeToo).
 Election influence tracking (e.g., how misinformation spreads).

D. Healthcare & Epidemiology

 Tracking disease spread in human populations.
 Understanding doctor-patient referral networks.

E. Cybersecurity & Fraud Detection

 Detecting fraud rings in financial networks.
 Identifying hacker communities and cyber threats.

3. Methods in Social Network Analysis

A. Graph-Based Methods
SNA uses graph theory to represent and analyze networks.
Degree Centrality → Number of direct connections a node has.
Betweenness Centrality → How often a node acts as a bridge between others.
Closeness Centrality → How close a node is to all others in the network.
Eigenvector Centrality → A measure of influence, used in Google's PageRank algorithm.
B. Community Detection Algorithms
Louvain Algorithm → Detects groups with strong internal connections.
Girvan-Newman Algorithm → Finds communities by removing important edges.
C. Network Visualization Techniques
Force-directed graphs (e.g., Gephi visualization).
Heatmaps and adjacency matrices for analyzing connections.
4. Tools for Social Network Analysis
A. Open-Source & Programming-Based Tools
 NetworkX (Python) → Powerful graph analysis tool.
 Gephi → Interactive network visualization software.
 igraph (R/Python) → Statistical network analysis.

B. AI & Machine Learning-Based Tools

 Node2Vec & DeepWalk → Machine learning for network embeddings.
 Graph Neural Networks (GNNs) → AI-based network predictions.

C. Social Media & Cybersecurity Tools

 Brandwatch, Hootsuite → Social media sentiment & influence analysis.
 Maltego → Cyber threat intelligence & fraud detection.

5. Challenges in Social Network Analysis

 Scalability → Large networks (e.g., Facebook, Twitter) require efficient algorithms.
 Data Privacy Issues → Ethical concerns about personal data analysis.
 Misinformation & Bias → AI-driven SNA can be manipulated by biased data.

1. Introduction to Social Network Analysis (SNA)

Social Network Analysis (SNA) is a research methodology used to study relationships, structures, and interactions
among individuals, groups, or organizations within a network. It applies graph theory, statistics, and computational
models to understand how information, influence, and behaviors spread in social structures.

Why is SNA Important?

Helps in understanding social relationships and influence patterns.
Identifies key influencers, communities, and network structures.
Widely used in sociology, marketing, cybersecurity, healthcare, and politics.

2. Historical Development of Social Network Analysis

A. Early Foundations (Pre-1900s – 1940s)
The roots of SNA trace back to sociology, anthropology, and psychology.

Sociometry (1930s) → Jacob Moreno developed sociometry to map and measure social relationships in small
groups.
Anthropological Studies → Researchers like Radcliffe-Brown and Malinowski studied kinship and tribal networks.
B. Structuralism & Graph Theory (1950s – 1970s)
Mathematical Foundations → Leonhard Euler's Graph Theory (1736) was later applied to social networks.
Small-World Phenomenon (1967) → Stanley Milgram introduced the “six degrees of separation” concept, showing
that people are connected by short chains.
Mark Granovetter (1973) → Introduced the “Strength of Weak Ties” theory, explaining how weak connections are
crucial for information flow.
C. Computational Advancements & Large-Scale Networks (1980s – 1990s)
Network Centrality Measures (Freeman, 1979) → Developed metrics like degree centrality, closeness centrality, and
betweenness centrality.
Sociologists like Duncan Watts & Steven Strogatz (1998) studied small-world networks, leading to breakthroughs in
understanding complex social structures.
D. Digital Era & Social Media Networks (2000s – Present)
Big Data & Machine Learning → AI-driven analysis of large-scale networks.
Rise of Online Social Networks → Facebook, Twitter, LinkedIn enabled real-time social network analysis.
Graph Neural Networks (GNNs) → Deep learning applied to SNA for recommendation systems, fraud detection,
and influence modeling.
3. Tools & Technologies Used in Social Network Analysis
A. Graph Theory-Based Tools
  Gephi → Visualization of network structures.
 NetworkX (Python) → Computational network analysis.
 igraph (Python/R) → Statistical analysis of networks.

B. AI & Machine Learning-Based Approaches

 Node2Vec & DeepWalk → Machine learning for network embeddings.
 Graph Neural Networks (GNNs) → AI for network predictions and classification.

C. Big Data & Social Media Analytics

 Brandwatch, Hootsuite → Social media influence tracking.
 Google Knowledge Graph → AI-powered semantic network analysis.