Download CSE 291-F: Graph Mining and Network Analysis

CSE291-F:GraphMiningandNetworkAnalysis
TentativeSyllabus
DepartmentofComputerScienceandEngineering,UCSanDiego,Spring2017
Instructor:FragkiskosMalliaros
Email:[email protected]
Office:AtkinsonHall,4111
Web:http://fragkiskos.me
CourseOverview
Networks (or graphs) have become ubiquitous as data from diverse disciplines can
naturally be mapped to graph structures. Socialnetworks, such as academic collaboration
networksandinteractionnetworksoveronlinesocialnetworkingapplicationsareusedto
representandmodelthesocialtiesamongindividuals.Informationnetworks,includingthe
hyperlink structure of the Web and blog networks, have become crucial mediums for
information dissemination, offering an effective way to represent content and navigate
through it. A plethora of technological networks, including the Internet, power grids,
telephonenetworksandroadnetworksareanimportantpartofeverydaylife.Theproblem
ofextractingmeaningfulinformationfromlargescalegraphdatainanefficientandeffective
way has become crucial and challenging with several important applications and towards
thisend,graphminingandanalysismethodsconstituteprominenttools.
Thegoalofthiscourseistopresentrecentandstate-of-the-artmethodsandalgorithmsfor
exploring,analyzingandmininglarge-scalenetworks,aswellastheirpracticalapplications
invariousdomains(e.g.,socialscience,theweb,biology).
LearningObjectives
This course aims to introduce students to the field of mining social and information
networksby(i)coveringawiderangeoftopics,methodologiesandrelatedapplications;(ii)
giving the students the opportunity to obtain hands-on experience on dealing with graph
dataandgraphminingtasksthroughexercisesandprojects.
Weexpectthatbytheendofthecourse,thestudentswillhaveathoroughunderstandingof
various graph mining and learning tasks, will be able to analyze large-scale graph data as
wellastoformulateandsolveproblemsthatinvolvegraphstructures.
Prerequisites
Thereisnoofficialprerequisiteforthiscourse.However,thestudentsareexpectedto:(i)
havebasicknowledgeofgraphtheoryandlinearalgebra;(ii)befamiliarwithfundamental
data mining and machine learning tasks; (iii) be familiar with at least one programming
language(e.g.,Pythonoranylanguageoftheirpreference).
CourseSchedule(Tentative)
Week
Topics
1
•
•
Introductionandoverviewofgraphmining
Graphtheoryandlinearalgebrarecap
2
•
Real-worldnetworkproperties(nodedegreedistribution,shortest
paths,clusteringcoefficient,smallworldphenomenon,propertiesof
timeevolvinggraphs)
Networkgenerativemodels(randomgraphs,preferential
attachmentmodel,Kroneckergraphs,stochasticblockmodels)
Randomwalks
Graphcentrality
Linkanalysisalgorithms(PageRankandHITS)
Graphclusteringandcommunitydetection(spectralclustering,
graphpartitioning,modularity-basedalgorithms,community
structureofreal-worldgraphs,overlappingcommunities)
Nodesimilarity
Linkprediction
Graphkernelsandgraphsimilarity
Graphclassification
Learningembeddingsingraphs
Applicationsinnodeclassificationandlinkprediction
Influentialspreaders
Influencemaximization
Topicsingraphminingandapplications
Densesubgraphdetection
Richnetworkstructures(signedandmultilayernetworks)and
applications
Anomalydetection
Geo-socialandlocationbasednetworks
Projectpresentations
•
3
4
5
6
7
8
9
10
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Evaluation(Tentative)
Theevaluationofthecoursewillbebasedonthefollowing:
1. Hands-on practical assignments: the hands-on assignments (two assignments)
will familiarize the students with basic graph mining and analysis tasks (e.g.,
analysis of a social or information network such as the DBLP co-authorship
network;crawlthedataandconstructthegraph;examinethestructuralproperties
ofthegraph;findtheunderlyingcommunities).
2. Researchproject:thiswillbethemaincomponentfortheevaluationofthecourse.
The students are expected to form groups of 3-4 people, propose a topic for their
project,andfinally,presenttheprojectattheendoftheclass(eitherinclassorin
postersessionthatwewillorganize).
ReadingMaterial
Mostofthematerialofthecourseisbasedonresearcharticles.Someofthetopicsarealso
coveredbythefollowingbooks(allofthemarepubliclyavailable):
• DavidEasleyandJonKleinberg.Networks,Crowds,andMarkets.Cambridge
UniversityPress,2010.
• DeepayanChakrabartiandChristosFaloutsos.GraphMining:Laws,Tools,andCase
Studies.SynthesisLecturesonDataMiningandKnowledgeDiscovery,Morgan&
ClaypoolPublishers,2012.
• Albert-LászlóBarabási.NetworkScience.CambridgeUniversityPress,2016.