NPTEL Social Networks Week 2 Assignment Answers 2025
1. Which of the following could be used as an indicator of a central ingredient in the Ingredients Network?
- High degree centrality
- Low clustering coefficient
- High betweenness centrality
- Minimum spanning tree membership
Answer : See Answers
2. If the network is unweighted and undirected, which metric would best indicate an ingredient that connects various ingredient clusters?
- Degree centrality
- Betweenness centrality
- PageRank
- Closeness centrality
Answer : See Answers
3. Which of the following actions might increase the clustering coefficient of a node representing an ingredient?
- Adding edges among its neighbors
- Using it in recipes that already share many other ingredients
- Decreasing its degree
- Connecting it to an isolated node
Answer : See Answers
4. What is the best way to identify substitute ingredients in the network?
- Using PageRank
- Using community detection to find clusters of similar-use ingredients
- Finding the node with maximum degree
- Running BFS from each node
Answer : See Answers
5. Ingredients with high clustering coefficient are likely to:
- Appear in tightly-knit groups of recipes
- Be used rarely
- Be part of cultural or cuisine-specific subgroups
- Form bridges between disparate recipes
Answer : See Answers
6. The network of ingredients is most similar in structure to which type of network?
- Tree
- Star
- Small-world network
- Linear chain
Answer : See Answers
The Synonymy Network
A team of linguists constructs a network where each node is a word, and edges connect words that are synonyms. They analyze how meanings evolve over time and find that certain hubs like “big”, “good”, or “fast” have dozens of connections, while niche terms like “gargantuan” or “stellar” sit on the periphery.
7. What graph property makes a word like “good” a hub?
- High degree
- High centrality
- Low clustering coefficient
- High eigenvector falloff
Answer : See Answers
8. What type of network model is most appropriate for modeling the synonymy network?
- Tree
- Undirected unweighted graph
- Directed weighted graph
- Bipartite graph
Answer : See Answers
9. Why might peripheral nodes in the synonym network be important?
- They add richness to language use
- They might connect rarely used synonym groups
- They have highest centrality
- They always form cycles
Answer : See Answers
10. What kind of analysis can help group together words with similar meanings?
- DFS traversal
- Community detection
- Shortest path algorithm
- Bipartite projection
Answer : See Answers
11. In the synonymy network, a node with low degree but high betweenness might indicate:
- A bridge between two different meaning clusters
- A context-specific synonym
- A very frequent word
- An isolated component
Answer :
12. What type of network metric can help identify formal vs informal word usage patterns?
- PageRank
- Clustering within sociolect subgraphs
- Tree depth
- Diameter of network
Answer :
The Web Graph
The structure of the web is modeled as a directed graph, where nodes are webpages and edges represent hyperlinks. Researchers analyzing the web graph discover that certain sites like Wikipedia and government portals have high PageRank, while some newer sites have low visibility.
13. PageRank is influenced by:
- Number of incoming links
- Quality (PageRank) of linking pages
- Number of outgoing links
- Number of total users
Answer :
14. A page with many incoming links from low-ranked pages is likely to:
- Have very high PageRank
- Never rank at all
- Have modest PageRank depending on damping factor
- Break the algorithm
Answer :
15. In the web graph, a dangling node is:
- A node with no outgoing links
- A node with no incoming links
- A broken hyperlink
- A self-looped node
Answer :
16. Which graph algorithm is most appropriate to identify top influencers in the web graph?
- Dijkstra’s algorithm
- PageRank
- BFS
- Kruskal’s algorithm
Answer : See Answers
17. Which of these changes could improve a site’s PageRank?
- Getting linked from authoritative websites
- Reducing number of outbound links on important pages
- Removing internal links
- Adding self-loops
Answer :
18. The Web Graph most resembles:
- Undirected graph
- Balanced binary tree
- Bipartite graph
- Directed scale-free network
Answer :
Social Network Datasets – Analysis using Graphs
You are working with a dataset of a social media platform. Each user is a node, and edges represent friend connections. Using Python’s NetworkX, you aim to understand central influencers, isolate cliques, and simulate potential viral content spread.
19. Which Python libraries are best suited for analyzing such social network graphs?
- NetworkX
- Pandas (for data preprocessing)
- TensorFlow
- Seaborn
Answer :
20. What metric would best identify a user who connects disparate groups?
- Closeness centrality
- Betweenness centrality
- Degree centrality
- Eigenvector centrality
Answer :
21. Community detection can help:
- Identify social circles
- Predict content spread boundaries
- Compute path lengths
- Rank friends
Answer :
22. A tightly-knit group of users with many mutual friends is likely to form:
- A PageRank cluster
- A path component
- A community
- A bipartite set
Answer :
23. Which of these plots would be helpful in visualizing the degree distribution?
- Histogram
- Log-log plot
- Heatmap
- Confusion matrix
Answer :
24. What does a power-law distribution in degree typically suggest?
- Uniform connectivity
- Presence of hubs
- Graph is bipartite
- Random structure
Answer :
Emergence of Connectedness
As a new mobile messaging app gains popularity, its user network grows. Initially, users are isolated, but as invitations spread and group chats are created, a giant connected component emerges. Researchers study when this happens and how rapidly such transitions occur.
25. The sudden emergence of a giant connected component is an example of:
- Phase transition in networks
- Percolation theory in action
- Clique formation
- Graph diameter reduction
Answer : See Answers
26. As new edges are added randomly, when is a giant component most likely to emerge?
- After adding 10% of edges
- Near the critical threshold in Erdős–Rényi model
- Immediately
- When all nodes have same degree
Answer :
27. Factors influencing the rate of connectedness emergence include:
- Average degree
- Number of nodes
- Time of day
- Color of nodes
Answer :
28. Which real-world phenomenon reflects the same principle?
- A virus mutating
- A trending hashtag
- A viral video suddenly reaching millions
- Blockchain mining
Answer :
29. Which of these methods can detect when a giant component has formed?
- Analyzing component sizes over time
- Checking network diameter
- BFS from random node
- Comparing node colors
Answer :
30. In graph theory, a component is:
- A central node
- A directed edge
- A maximal connected subgraph
- A subgraph with cycles only
Answer : See Answers
Online Learning Platforms – Collaboration and Influence
In an online learning platform, users interact by asking and answering questions, collaborating on assignments, and sharing resources. Each user is a node, and interactions such as replies, comments, or co-enrollments are edges. Analysts are interested in understanding influence dynamics, collaborative learning patterns, and knowledge diffusion.
31. Which of the following network properties would best identify influential contributors?
- High in-degree (number of incoming replies/comments)
- High eigenvector centrality
- Low clustering coefficient
- Membership in isolated components
Answer : See Answers
32. Which metric would help detect tightly-knit learning groups?
- Degree centrality
- Clustering coefficient
- PageRank
- Edge betweenness
Answer : See Answers
33. A user with high betweenness centrality is likely to:
- Connect different learning communities
- Facilitate knowledge transfer between groups
- Be isolated from core discussions
- Have the most followers
Answer : See Answers
34. Which algorithm is most suitable for identifying natural study groups in the network?
- BFS
- Dijkstra’s algorithm
- Community detection (e.g., Girvan–Newman or Louvain)
- Kruskal’s algorithm
Answer : See Answers
35. Factors contributing to strong collaborative ties may include:
- Frequent co-participation in threads
- Repeated peer review or feedback
- Random logins
- Degree of anonymity
Answer : See Answers
36. In such a learning network, a node with high closeness centrality can:
- Disseminate information quickly to the whole network
- Only influence its direct neighbors
- Be easily removed without effect
- Belong to disconnected components
Answer : See Answers
