JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD

Course Objectives

The objective of the course is to study the techniques to model, analyze, and

understand large-scale social media along with dynamic processes over social and information networks, and understand the link between qualitative and quantitative methods of social media mining.

Course Outcomes:

• 1. Understand the visualization of complex social networks
• 2. Analyze how to detect communities and influencers
• 3. Implementing K-means clustering
• 4. Implement classification with network information.

List of Experiments: Implement the following exercises using NodeXL / Python:

• 1. Collect and visualize complex social networks from Twitter, Wikipedia.
• 2. Compute vertex metrics and network metrics for the social graphs. (i)Degree centrality (ii)Eigenvector centrality (iii)Betweenness centrality (iv)PageRank
• (v) Closeness centrality (vi) Group centrality (vii) clustering coefficient
• 3. Visualize the social graphs reflecting the metrics.
• 4. Detect Bridge in a social graphs.
• 5. Detect communities and influencers: Brute-Force Clique Identification on Enron email data.
• 6. Girvan-Newman Algorithm for community detection.
• 7. Perform classification with network information: weighted vote relational neighbor classifier for Twitter data.
• 8. Perform sentiment analysis on IMDb dataset.
• 9. Apply k-means clustering algorithm on IMDb dataset.
• 10. Apply user-based collaborative filtering on Amazon review data/Amazon product data.
• 11. Apply item-based collaborative filtering on Amazon review data/Amazon product data.
• 12. Predict individual behavior of users in social media.

TEXT BOOK:

• 1. Social Media Mining (An Introduction), Reza Zafarani, Mohammad Ali Abbasi, Huan Liu, Cambridge University Press, Draft Version: April 20, 2014

REFERENCES:

• 1. Mining the Social Web, 2nd Edition Data Mining Face book, Twitter, LinkedIn, Google+, GitHub, and More, Matthew A. Russell Publisher: O'Reilly Media.
• 2. Social Media Mining with R [Kindle Edition] Nathan Danneman Richard Heimann.