CPSC 534L: Topics in Data Management & Mining  Social and Information Networks   

                                                                    Laks V.S. Lakshmanan

Classes: TR 9:30--11 am, DMP 101, Sept.--Nov. 2018.

Sneak Preview on Tuesday, Sept. 4 at the CSGSA Orientation, Course Pitch.

First lecture: Thursday, Sept. 6.

Watch for updates.

Quick Jumps:

Past Student Talks 
Marking Scheme
Instructions and Tips on Paper Presentations 
Guidelines for Speakers and Discussion Leaders 
Evaluation of Presentations and Discussion Leadership 
Past Project Topics 
Project Deadlines and Deliverables 
Project Talk Scheule

Introduction: Recent years have witnessed a tremendous interest in networked information structures. Examples of such strctures abound and span the spectrum, from one extreme,   online social networks, to homogeneous and heterogeneous information networks consisting, e.g., of people, resources, and their artifacts, to networks of informational and computational resources, to biological networks and  to brain networks.  Conducting analysis and mining on very large networks is an important part of harnessing Big Data. In an apparently different trend, recommender and collaborative tagging/rating/reviewing systems have gained tremendous popularity. In particular, there is growing evidence suggesting that recommendations will serve as a compelling and complementary alternative to search over large collections of information. Indeed, an integrated paradigm that combines search with recommendations will significantly boost the quality of resource discovery. 

In this course, taking social networks as a representative example of networked information structures, we will focus on key data mining and computational problems in Social Networks and Recommender Systems. Our scope will include the technical aspects of modeling, searching, and mining motivated by challenging research problems that arise in their context and on designing and analyzing algorithms for solving those problems. 

Here is an Outline of the course, the place to go to for all lecture slides. We will be using Piazza for all online discussion related to the course as well as for managing all submissions, announcements, and class resources.Sign up here. See CPSC534LHOME for a detailed description of the class format and related information.

Click here for the Intro slides. 


Outline is tentative and is based on a past offering. Currently being revamped.

1''Introduction''Origins, early history, sociologist’s perspective, social capital, centrality, social web, web 2.0 and web 3.0, focus of this course.
2Structural Analysis of NetworksLink Analysis  (Overview)

Team Formation
What do links say about importance of pages?

How best to form teams to accomplish tasks?
Ad hoc Community search (cocktail party planning)

Can we divide up the n/w along lines of "interest"?
Community Detection
Community models: k-cores and k-truss

Community Search
How do you query communities?
Viral MarketingInfluence in Social Networks -- Intro.
Influence Propagation
General remarks and applications.
Whether, when and how does influence propagate in a n/w?
Equivalence of LT & Live Edge ModelsWhy exactly is LT submodular, again?
Learning ModelsWhere do the influence probabilities come from?
Non-competitive VM Parting Shots
Competitive VM
Competitive VM -- A Host's Perspective
Social Advertising.
Handling Competition.
Recommender SystemsRecommender Systems -- Intro.; Content-based vs.Collaborative FilteringRecommend items based on their intrinsic properties or on the "wisdom" of the crowd?
Memory-based vs. Model-basedUse memory of past user behavior as is or build models?
Top-K Search and Package Recommendations & Other Novel Recommendation Problems
Is maximizing prediction accuracy of movies/songs the only goal?



Marking Scheme: