computational discovery in evolving complex networks

Yongqin Gao Dissertation DefenseDecember 2006

Computational Discovery in Evolving Complex Networks

Yongqin Gao

Advisor: Greg Madey

Yongqin Gao December 2006

Dissertation Defense

Outline• Background• Methodology for Computational Discovery• Problem Domain – OSS Research• Process I: Data Mining• Process II: Network Analysis• Process III: Computer Simulation• Process IV: Research Collaboratory• Contributions• Conclusion and Future Work



Background• Network research gains more attentions

– Internet

– Communication network

– Social network

– Software developer network

– Biological network

• Understanding the evolving complex network– Goal I: Search

– Goal II: Prediction

• Computational scientific discovery



Computational DiscoveryOur Methodology

ResearchCollaboratory

Data Mining

NetworkAnalysis

ComputerSimulation

Discovery Assessment

RevisionFeedback

Researcher

Community Members

Contribution Reference

Initialization



Problem Domain• Open Source Software Movement

– What is OSS• Free to use, modify and distribute and source code available

and modifiable

• Potential advantages over commercial software: Potentially high quality; Fast development; Low cost

– Why study OSS (Goal)• Software engineering — new development and coordination

methods

• Open content — model for other forms of open, shared collaboration

• Complexity — successful example of self-organization/emergence



Glory of OSSNumber of Active Apache Hosts



Problem Domain• SourceForge.net community

– The biggest OSS development communities– 134,751 registered projects– 1,439,773 registered users



Problem Domain• Our Data Set

– 25 monthly dumps since January 2003.– Totally 460G and growing at 25G/month.– Every dump has about 100 tables.– Largest table has up to 30 million records.

• Experiment Environment– Dual Xeon 3.06GHz, 4G memory, 2T storage– Linux 2.4.21-40.ELsmp with PostgreSQL 8.1



Related Research• OSS research

– W. Scacchi, “Free/open source software development practices in the computer game community”, IEEE Software, 2004.

– C. Kevin, A. Hala and H. James, “Defining open source software project success”, 24th International Conference on Information Systems, Seattle, 2003.

• Complex networks– L.A. Adamic and B.A. Huberman, “Scaling behavior of

the world wide web”, Science, 2000.– M.E.J. Newman, “Clustering and preferential attachme

nt in growing networks”, Physics Review, 2001.



Process I: Data Mining• Related Research:

– S. Chawla, B. Arunasalam and J. Davis, “Mining open source software (OSS) data using association rules network”, PAKDD, 2003.

– D. Kempe, J. Kleinberg and E. Tardos, “Maximizing the spread of influence through a social network”, SIGKDD, 2003.

– C. Jensen and W. Scacchi, “Data mining for software process discovery in open source software development communities”, Workshop on Mining Software Repositories, 2004.



Process I: Data Mining

Raw data

Relevant data

Data Purging

Feature Selection

Algorithm Application

Data Preparation

Database



Process I: Data Mining• Data Preparation

– Data discovery• Locating the information

– Data characterization • Activity features: user categorization• Network features

– Data assembly• Data Purging

– Treatment about data inconsistency• Unifying the date presentation by loading into single depository

– Treatment about data pollution• Removing “inactive” projects

• Feature Selection– This method is used to remove dependent or insignificant features.– NMF (Non-negative Matrix Factorization)



Process I: Data Mining• Result I

– Significant features• By feature selection, we can identify the significant

feature set describing the projects.

• Activity features: “file_releases”, “followup_msg”, “support_assigned”, “feature_assigned” and task related features

• Network features: “degrees”, “betweenness” and “closeness”



Process I: Data Mining• Distribution-based clustering (Christley, 2005)

– Clustering according to the distribution of features instead of values of individual feature

– We assume every entity (project) has an underlying distribution of the feature set (activity features)

– Using statistical hypothesis test• Non-parametric test• Fisher’s contingency-table test is used

– Joachim Krauth, “Distribution-free statistics: an application-oriented approach”, Elsevier Science Publisher, 1988.



Process I: Data Mining• Procedure:

While (still unclustered entities)Put all unclustered entities into one clusterWhile (some entities not yet pairwise compared)

A = Pick entity from clusterFor each other entity, B, in cluster not

yet compared to ARun statistical test on A and BIf significant result

Remove B from cluster

• Worst case complexity: O(n2)



Process I: Data Mining• Result II• Unsupervised learning

– Distribution-based method used to cluster the project history using the activity distribution

– We named the clusters using ID and the results are shown in the table

– High support and confidence in evaluation

Cluster ID Size

1 89709

2 9191

3 2060

Total 100960



Process I: Data Mining• Two sample

distributions from different categories

• Unbalanced feature distribution → could be “unpopular”

• Balanced feature distribution → could be “popular”

20

1641

3488

22 0

312

736

229

1510

534

82 12128 0 4

0

500

1000

1500

2000

2500

3000

3500

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Activity Category

Cluster 1

134

3781

8435

4310

21792537

667

9169

7134

601

2411

1651

0399

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Activity Category

Cluster 3



Process I: Data Mining• Discoveries in Process I

– Significant feature set selection• Network features are important

• Further inspection in next process

– Distribution based predictor• Based on the activity feature distribution

• Prediction of the “popularity” based on the balance of the activity feature distribution

• Benefit of these discoveries– For collaboration based communities, these discoveries

can help in resource allocation optimization.



Process II: Network Analysis• Why network analysis

– Assess the importance of the network measures to the whole network and to individual entity in the network

– Inspect the developing patterns of these network measures

• Network analysis– Structure analysis– Centrality analysis– Path analysis



Process II: Network Analysis• Related research:

– P. Erdös and A. Rényi, “On random graphs”, Publicationes Mathematicae, 1959.

– D.J. Watts and S. H. Strogatz, “Collective dynamics of small-world networks”, Nature, 1998.

– R. Albert and A.L. Barabάsi, “Emergence of scaling in random networks”, Science, 1999.

– Y. Gao, “Topology and evolution of the open source software community”, Master Thesis, 2003.



Process II: Network Analysis• Structure Analysis

– Understanding the influence of the network structure to individual entities in the network

– Inspected measures• Approximate diameter

• Approximate clustering coefficient

• Component distribution

1)/log(

)/log(

12

1 zz

zND

)32())((

1

1

32111

21212

C



Process II: Network Analysis• Conversion among C-NET, P-NET and D-

NET



Process II: Network Analysis• Result I

– Approximate Diameters• D-NET: between (5,7) while network size ranged

from 151,803 to 195,744.

• P-NET: between (6,8) while network size ranged from 123,192 to 161,798.

– Approximate Clustering Coefficient• D-NET: between (0.85, 0.95)

• P-NET: between (0.65, 0.75)



Process II: Network Analysis• Result I



Process II: Network Analysis• Centrality Analysis

– Understanding the importance of individual entities to the global network structure

– Inspected measures:• Average Degrees

• Degree Distributions

• Betweenness

• Closeness

Vtvs st

st vvB

)(

)(

Vt G tvd

vC),(

1)(



Process II: Network Analysis• Result II

– Average Degrees• Developer degree in C-NET: 1.4525

• Project degree in C-NET: 1.7572

• Developer degree in D-NET: 12.3100

• Project degree in P-NET: 3.8059



Process II: Network Analysis• Result II (Degree distributions in C-NET)



Process II: Network Analysis• Result II (Degree distributions in D-NET

and P-NET)



Process II: Network Analysis• Result II

– Average Betweenness• P-NET: 0.2669e-003

– Average Closeness• P-NET: 0.4143e-005

– Normally these two measures yield very small value in large networks (N>10,000).



Process II: Network Analysis• Path Analysis

– Understanding the developing patterns of the network structure and individual entities in the network

– Inspected measures:• Active Developer Percentage• Average Degrees• Diameters• Clustering coefficients• Betweenness• Closeness



Process II: Network Analysis• Result III (Active entities)



Process II: Network Analysis• Result III (Average degrees in C-NET)



Process II: Network Analysis• Result III (Average degrees in D-NET and

P-NET)



Process II: Network Analysis• Result III (Diameters in D-NET and P-

NET)



Process II: Network Analysis• Result III (Clustering coefficients for D-

NET and P-NET)



Process II: Network Analysis• Result III (Average betweenness and closen

ess for P-NET)



Process II: Network AnalysisMeasures D-NET P-NET C-NET

Average Degree Yes Yes Yes

Diameter Yes Yes N/A

Clustering Coefficient Yes Yes N/A

Degree Distribution Yes Yes Yes

Component Distribution N/A Yes N/A

Major Component N/A Yes N/A

Average Betweenness Yes Yes N/A

Average Closeness Yes Yes N/A

Active Entity Size Development Yes Yes Yes

Average Degree Development Yes Yes Yes

Diameter Development Yes Yes N/A

Clustering Coefficient Development Yes Yes N/A

Average Betweenness Development Yes Yes N/A

Average Closeness Development Yes Yes N/A



Process II: Network Analysis• Discoveries in Process II:

– Measures of structure analysis and centrality analysis all indicate very high connectivity of the network.

– Measures of path analysis reveal the developing patterns of these measures (life cycle behavior).

• Benefits of these discoveries– High connectivity in a network is an important feature

for information propagation, failure proof. Understanding this discovery can help us improve our practices in collaboration networks and communication networks.

– Understanding the developing patterns of these network measures provides us a method to monitor network development and to improve the network if necessary.



Process III: Computer Simulation

• Related Research:– P.J. Kiviat, “Simulation, technology, and the decision p

rocess”, ACM Transactions on Modeling and Computer Simulation,1991.

– R. Albert and A.L. Barabási, “Emergence of scaling in random networks”, Science, 1999.

– J. Epstein R. Axtell, R. Axelrod and M. Cohen, “Aligning simulation models: A case study and results”, Computational and Mathematical Organization Theory, 1996.

– Y. Gao, “Topology and evolution of the open source software community”, Master Thesis, 2003.




• Iterative simulation method– Empirical dataset

– Model

– Simulation

• Verification and validation– More measures

– More methods

Model

SimulationEmpirical

DataCollection

Verification

Validation




• Previous iterated models (master thesis):– Adapted ER Model– BA Model– BA Model with fitness– BA Model with dynamic fitness

• Iterated models in this study– Improved Model Four (Model I)– Constant user energy (Model II)– Dynamic user energy (Model III)




• Model I– Realistic stochastic procedures.

• New developer every time step based on Poisson distribution

• Initial fitness based on log-normal distribution

– Updated procedure for the weighted project pool (for preferential selection of projects).




• Average degrees




• Diameter and CC




• Betweenness and Closeness




• Degree Distributions




• Deficit in the measures




• Model II– New addition: user energy.– User energy

• the “fitness” parameter for the user

• Every time a new user is created, a energy level is randomly generated for the user

• Energy level will be used to decide whether a user will take a action or not during every time step.




• Degree distributions for Model II




• Deficit in the measures




• Model III– New addition: dynamic user energy.– Dynamic user energy

• Decaying with respect to time

• Self-adjustable according to the roles the user is taking in various projects.




• Degree distributions (Model III)



Process III: Computer SimulationModels Measures Patterns in Data Simulated Patterns

Model I

(more realistic distributions)

Developer Distribution Power Law (large tail) Power Law (small tail)

Project Distribution Power Law (small tail) Power Law (large tail)

Average Degrees Increasing Increasing

Clustering Coefficient Decreasing Decreasing

Diameter Decreasing Decreasing

Average Betweenness Decreasing Decreasing

Average Closeness Decreasing Decreasing

Model II

(constant user energy)

Developer Distribution Power Law (large tail) Power Law (large tail)

Project Distribution Power Law (small tail) Power Law (reasonable tail)






Model III

(dynamic user energy)

Developer Distribution Power Law (large tail) Power Law (large tail)

Project Distribution Power Law (small tail) Power Law (small tail)









• Discoveries in Process III– Expanding the network models for modeling ev

olving complex networks (more parameters)– Providing a validated model to simulate the co

mmunity network at SourceForge.net

• Benefits of these discoveries– Expanded network models can benefit other res

earchers in complex networks.– Validated model for SourceForge.net can be us

ed to study other OSS communities or similar collaboration networks.



Process IV: Research Collaboratory

• Related Research:– G. Chin Jr. and C. Lansing, “The biological scie

nces collaboratory”, Mathematics and Engineering Techniques in Medicine and Biological Sciences, 2004.

– L. Koukianakis, “A system for hybrid learning and hybrid psychology”, Cybernetics and Information Technologies, Systems and Applications, 2003.

– NCBI, FlyBase, Ensembl, VectorBase




• What is Collaboratory?– An elaborate collection of data, information,

analytical toolkits and communication technologies

– A new networked organizational form that also includes social processes, collaboration techniques and agreements on norms, principles, value, and rules




Data Repository

Wiki Interface

Query

Researchers

RPC Browse

Researchers

Presentation Tier

This top tier is the user interface.The main function of the interface isto translate tasks and results tosomething the user can understand.

Logic Tier

This tier coordinates the webinterface and the data storage,moves and processes data betweenthe two surrounding tiers.

Data Tier

Here information is stored andretrieved from a database. Theinformation will then be passed backto user through the logic tier.




• Data tier - schema design

SF0205

SF0103

SF0405SF0305

SF0605

SF0705SF0805

SF0505

Every schema is adatabase dump

from theSourceForge.net

Timeline




• Data tier - connection pool

TimelineConnection Pool

ConnectionAssigner

LogicTier

ConnectionRequest

PersistentLink

PersistentLink

PersistentLink




• Presentation Tier– Various access

methods

– Documentation and references

– Community support

– Wiki interface




• Logic Tier– Interactive web query system

• Authorized user can submit query to the back end repository through the web query

• Results are provided by files with various formats

– Dynamic web schema browser• Authorized user can access the dynamic schema of

the repository through the schema browser




• Utilization reports– Monthly statistics (June 2006)

• Total queries submitted: 16,947

• Total data files retrieved: 13,343

• Total bytes of query data downloaded: 26,684,556,278

• Programmable access method– Programmable access method should be provided

for complicated access– Web services planned




• Results in Process IV– Designing, implementing and maintaining a

research collaboratory for OSS related research.

• Benefits of these results– OSS researchers can access one of the most

complete data sets for a OSS community development.

– By providing the community service to OSS researchers, the collaboratory can help in sparkling, improving and promoting research ideas about OSS.



Contributions• Designed and demonstrated a computational discovery methodology

to study evolving complex networks using research on OSS as a representative problem domain

• Understanding the OSS movement by applying the methods.– Process I: data mining

• Identifying significant features to describe a project• Using distribution based clustering to generate a distribution based predictor to

predict the “popularity” of a project– Process II: network analysis

• Introducing more complete analysis to inspect more complete data set from SourceForge.net.

• Discovering high connectivity and possible life cycle behaviors in both the network structure and individuals in the network

– Process III: computer simulation• Introducing more parameters in modeling evolving complex networks• Generating a “fit” model to replicate the evolution of the SourceForge.net

community.– Process IV: research collaboratory

• Designing, implementing and maintaining a research collaboratory to host the SourceForge.net data set and provide community support for OSS related researches.



Publications to-date• Y. Gao; G. Madey and V. Freeh. “Modeling and simulation of the open so

urce software community”, ADSC, San Diego, 2005.• Y. Gao and G. Madey. “Project development analysis of the oss communit

y using st mining”, NAACSOS, Notre Dame, 2005.• S. Christley; Y. Gao; J: Xu and G. Madey. “Public goods theory of the op

en source software development community”, Agent, Chicago, 2004.• Y. Gao, Y. Huang and G. Madey, “Data Mining Project History in Open S

ource Software Communities”, NAACSOS, Pittsburgh, 2004.• J. Xu, Y. Gao, J. Goett and G. Madey, “A Multi-model Docking Experime

nt of Dynamic Social Network Simulations”, Agent, Chicago, 2003.• Y. Gao, V. Freeh, and G. Madey, “Analysis and Modeling of the Open So

urce Software Community”, NAACSOS, Pittsburgh, 2003. • Y. Gao, V. Freeh, and G. Madey, “Conceptual Framework for Agent-base

d Modeling and Simulation”, NAACSOS, Pittsburgh, 2003. • G. Madey; V. Freeh; R: Tynan and Y. Gao. “Agent-based modeling and si

mulation of collaborative social networks”, AMCIS, Tampa, 2003.• Y. Gao; V. Freeh and G. Madey. “Topology and evolution of the open sou

rce software community”, SwarmFest, Notre Dame, 2003.



Publication Plan• Chapter III (data mining)

– Journal of Machine Learning Research – Journal of Systems and Software

• Chapter IV (network analysis)– Journal of Network and Systems Management– Journal of Social Structure

• Chapter V (computer simulation)– Spring Simulation Conference 2007 (under review)– IEEE Computing in Science and Engineering

• Chapter VI (research collaboratory)– CITSA 2007– Journal of Computer Science and Applications



Conclusion and Future Work• Cyclic computational discovery method for

studying evolving complex networks• Study of Open Source Software by applying this

method• Future works:

– Maintaining and expanding the collaboratory– Verifying the discoveries in the SourceForge.net

against further accumulated database dump from SourceForge.net

– Applying our simulation model on other software development communities

– Extending our methodology to other evolving complex networks like Internet, communication network and various social networks



Acknowledgement• My advisor: Dr. Madey• My committee members:

– Dr. Flynn– Dr. Striegel– Dr. Wood

• My Colleagues: – Scott Christley, Yingping Huang, Tim Schoenharl, Matt Van Antw

erp, Ryan Kennedy, Alec Pawling and Jin Xu

• SourceForge.net managers:– Jeff Bates, VP of OSTG Inc.– Jay Seirmarco, GM of SourceForge.net.

• US NSF CISE/IIS-Digital Society & Technology, under Grant No. 0222829.



Questions



Case Study II

15850 dev[46]dev[83] 15850 dev[46]

dev[48]

15850 dev[46]dev[56]

15850 dev[46]dev[58]

6882 dev[58]dev[47]

6882 dev[47]dev[79]

6882 dev[47]dev[52]

6882 dev[47]dev[55]

7028 dev[46]dev[99]

7028 dev[46]dev[51]

7028 dev[46]dev[57]

7597 dev[46]dev[45]

7597 dev[46]dev[72]

7597 dev[46]dev[55]

7597 dev[46]dev[58]

7597 dev[46]dev[61]

7597 dev[46]dev[64]7597 dev[46]

dev[67]

7597 dev[46]dev[70]

9859 dev[46]dev[49]9859 dev[46]

dev[53]

9859 dev[46]dev[54]

9859 dev[46]dev[59]

dev[46]

dev[83] dev[56]

dev[48]

dev[52]

dev[79]

dev[72]

dev[51]

dev[57]

dev[55]

dev[99]

dev[47]

dev[58]

dev[53]

dev[58]

dev[65]

dev[45]

dev[70]

dev[67]

dev[59]

dev[54]

dev[49]

dev[64]

dev[61]

Project 6882

Project 9859

Project 7597

Project 7028

Project 15850

OSS Developer Network (Part)Developers are nodes / Projects are links

24 Developers5 Projects

2 hub Developers1 Cluster



Process I: Data Mining• Characteristics of data set

– Massive

– Incomplete, noisy, redundant

– Complex structures, unstructured

• Classic analysis tools are often inadequate and inefficient for analyzing these data, especially in exploratory research

• What is DM (Data mining)– Nontrivial extraction of implicit, previously unknown

and potentially useful information from data.



Process I: Data Mining• Feature Selection

– Given a non-negative n x m matrix V, find factors W (n, r) and H (r, m) , such that

V ≈ W *H– This is called the non-negative matrix

factorization (NMF) of the matrix V– NMF can be used on multivariate data to reduce

the dimension of the data set– By using NMF, we can reduce dimension from

m features to r features



Why NMF?• Feature extraction methods

– linear methods are simpler and more completely understood.

– nonlinear methods are more general and more difficult to analyze.

• Linear methods: – ICA: Independent Component Analysis– Matrix decomposition: PCA, SVD, NMF

• In practice, NMF is most popular and simple.• Dimensionality reduction is effective if the loss of

information due to mapping to a lower-dimensional space is less than the gain due simplifying the problem.



Process I: Data Mining• Feature-based Clustering

– Grouping data into K number of clusters based on features.

– The distance metrics used is Euclidean distance like

– Hierarchical K-Means is used.• The result is a binary tree.

• The root is the whole data set and the leaf clusters are the fine-grained clusters, which are the resulting K clusters.

n

iii yxED

0

2)(



Process I: Data Mining• Case Study Result II • Unsupervised learning

– K-Means method used to cluster the project history using the features we selected

– We named the clusters using ID and the results are shown in the table

– The result is not acceptable by evaluation

Cluster ID Size

1 6201

2 98

3 64824

4 2

5 4

6 29724

7 4

8 10

9 9

10 84

Total 100960




Admin_flags?

Administrator Core developer Co-developer Active user lurker

Grantcvs?

Yes

No

Yes

User_grouptable

artifacttable

Forumtable

People_jobtable

Project_tasktable

Doc_datatable

UNION

Othertables

User_project_acttable

Assigned?

Activities?

Yes

No

No

Yes

No



Clustering Result Evaluation• Evaluation test set generation

– Popular/unpopular projects

– Stratified sampling to make 500 projects

• Feature sets used– Popular feature set

– Activity Feature set (Page 34, Table 3.2)

– Network Feature set (Page35, Table 3.3)

• Generating rules for the test sets• Calculating the support and confidence value



Popularity Definition

Feature DescriptionDevelopers Number of core developers

Downloads Number of downloads

Site_views Number of views of the website

Subdomain_views Number of views of the subdomain

Page_views Number of views of the pages



Why K-MEAN?• The algorithm has remained extremely popular because it c

onverges extremely quickly in practice. In fact, many have observed that the number of iterations is typically much less than the number of points.

• K-Means is most successful algorithm in large data set (size>1000, dimension > 2) than GA and Evolution

• CLIQUE is sensitive to noise• CURE is not scalable O(n2logn)• CLARANS & BIRCH are not good for high dimension dat

a

• D. Arthur, S. Vassilvitskii (2006): "How Slow is the k-means Method?," Proceedings of the 2006 Symposium on Computational Geometry (SoCG).



K-MEAN• It maximizes inter-cluster (or minimizes

intra-cluster) variance, but does not ensure that the result has a global minimum of variance. Multiple run is needed.

• Elbow criterion



Distribution CategoriesCategory Feature

1 File release

2 New message

3 Followup message

4 Artifact request

5 Todo request

6 Support request

7 Feature request

8 Patch request

9 Bug reports

10 Bug assigned

11 Patch assigned

12 Feature assigned

13 Support assigned

14 Todo assigned

15 Artifact assigned



Process III: Computer SimulationStart

Stop

End of Simu?

WeightedProject Pool

User Action

No

Yes

Project ListUser List

Project PoolUpdate

JoinCreateIdle Drop

User_ProjectLinks

New UsersSimulation model

procedure




• Poisson Process:– It expresses the probability of a number of events

occurring in a fixed period of time if these events occur with a known average rate, and are independent of the time since the last event.

– PDF:

!);(

k

ekF

k




• Log-normal distribution:




• Kolmogorov-Smirnov test– Used to determine whether two underlying one-dimen

sional distributions differ.

– Two one-sided K-S test statistics are given by

))()(max(

))()(max(

xFxFD

xFxFD

nn

nn



Similar Publications• Chapter III (data mining)

– JMLR: G. Hamerly, E. Perelman..Using machine learning to guide simulation (Feb. 2006)

– JSS: S. Kim, J. Yoon..Shape-based retrieval in time-series database (Feb. 2006)

• Chapter IV (network analysis)– JNSM: Special Issue Self-Managing Systems and Networks – JoSS: The Journal of Social Structure (JoSS) is an electronic journal of th

e International Network for Social Network Analysis (INSNA) • Chapter V (computer simulation)

– SSC 2007: simulation co– IEEE/CSE: E. Luijten..Fluid simulation with monte carlo algorithm (2006

Vol. 8, Issue 2)• Chapter VI (research collaboratory)

– CITSA 2007: L. Koukianakis..A system for hybrid learning and hybrid psychology (2005)

– JCSA: S. Chen, K. Wen..An Integrated System for Cancer-Related Genes Mining from Biomedical Literatures (2006)

computational discovery in evolving complex networks

Documents