wwsss intro2016-final

Steffen Staab Web Science 1Institute for Web Science and Technologies · University of Koblenz-Landau, GermanyWeb and Internet Science Group · ECS · University of Southampton, UK &

Introduction to Web Science

Steffen StaabUniversity of Southampton &Universität Koblenz-Landau

Steffen Staab Web Science 2

Welcome in Koblenz!


9.00 am 10.30 am 3.00 pm

Thu Introduction to Web Science

Noshir Contractor Tutorial Poster session

Fri Internet and Law Nikolaus Forgo Tutorial Project work

Sat Web and Politics Stéphane Bazan Tutorial Project work

Mon Social Machines Jim Hendler Tutorial Project work

Tue Web Entrepreneurship Simon Köhl Tutorial Project work

Wed Computational Social Science

Nuria Oliver Tutorial Project presentation

Program Overview

Details at http://wwsss16.webscience.org/schedule


• Work independently and self-organised• Get feedback from your advisor • Choose your working space:

– D 238, D 239, A 308, B 005 on weekdays– E 412, E 413, E 414, A 308, B 005 on Saturday– or anywhere you like

• Final presentation on Wednesday afternoon (10 minutes per team)

Project work

Details at http://wwsss16.webscience.org/schedule


Social EventsThu Cocktails at city beach 6.00 pm Meet at main campus entry

Fri SommerUni party 8.00 pm Campus

Sat Free time Old town, KaleidosKOp festival, fortress, beer gardens, ...

Sun Excursion (few seats available!) 10.45 am Meet at main campus entry

Quarter final France vs. Island 9.00 pm Campus

Mon Free time Old town, fortress, beer gardens, ...

Tue Barbecue 7 pm Campus

SommerUni party 9 pm Campus

Wed SommerUni barbecue 6 pm Campus


Thursday

Noshir Contractor:Leveraging Web/Internet/Network Sciences (WINS) to address Grand Societal Challenges

9.00 am

10.30 amSteffen Staab and Jérôme Kunegis: Introduction to Web Science

3.00 pmPoster session6.00 pm

Cocktails at city beachmeeting point: main campus entrance

Breaks:• 10 am, • 12 am, • 2.30 pm


Friday

Nikolaus Forgó:Privacy Law and the Web: A Story of Love and Hate?

9.00 am

10.30 amRüdiger Grimm: Internet and Law

3.00 pmSupervised project work8.00 pm

SommerUni partyAt the campus – impossible to miss!

Breaks:• 10 am, • 12 am, • 2.30 pm


Thanks to our sponsors!


Let‘s get started!


Produce

Consume

Cognition

EmotionBehavior

SocialisationKnowledge

Observable Micro-

interactions in the Web

AppsProtocols

Data & InformationGovernance

WWW

Observable Macro-

effects in the Web

Web Science


Tertium Datur or Where Dijkstra was wrong

Computer Science

Science about

Computers

Astronomy

Telescope Science

Web Science

Science about the

Web

What is Web Science?


Web science is an emerging interdisciplinary field concerned with the study of large-scale socio-technical systems, such as the World Wide Web. It considers the relationship between people and technology, the ways that society and technology co-constitute one another and the impact of this co-constitution on broader society.

Wikipedia, 2016-06-29

Definition of Web Science


Agenda• What is Web Science?• What is the Web?

– Aspects of the Web at Large• How to investigate the Web?

– Observing the Web– An example using the architecture: bias in the Web– How to model aspects of the Web

• What is the past and the future of the Web?


I try to• classify, describe commonalities, find generalizations

I will not• be 100% correct, 100% complete

BUT• Please ask, suggest, complement,...WHENEVER you feel like it!DON‘T HESITATE ASKING!

A Note


What is the Web?


What is the Web: Aspects of the Web at LargeArchitecture methodology:• Draw pictures in different dimensions• Pictures do not compete with, but complement each

other

Not only technical pictures!


The Web as a Device

Software• Browsers

– IE, Firefox, Chrome,...• Web Servers

– Apache, Tomcat,..• Content Management

and Data Delivery– Wordpress, drupal,

databases...• Search Engines• ...

Standards• Uniform Resource

Locator (URL)• HyperText Transfer

Protocol (http)• HyperText Markup

Language (html)• Domain name service • ...+ many more


The Web as Content

For human consumpton (primarily)

Text, HypertextImagesVideoAudioMultimediaInteraction (Games...)BrailleMathematics

For machine consumption (primarily)MetadataDataOntologies


The Web as Content

Free• Informational• Advertisements• Goods & services• Web 2.0

– Wikipedia– facebook

Paid• Individual payments• Subscription• Micropayment

Freemium


The Web and its Stakeholders

People• Citizens• Customers• Leisure seekers• Workers• Software developers

Internet providers• Landline• Mobile• Nested providers

(internet cafe...)• Website hosts• Peer2peer networks (Bittorrent...)

Platform operators• Shops• News• Web 2.0• Payment• Advertisement networks• Trust centers

Government• Police• Military• Secret service• Law• Citizen services• Administration• Politics


The Web as a Process

Governing• Standards processes

– W3C– Internet Engineering

Task Force (IETF)– RFC

• Domain name registration

• Internet routing– E.g. „great Chinese

firewall“

Regulation• Legal

– copyright • where enforced

– hate speech– ...

• Private– E.g. Facebook,

Instagram ... Pictures, nipple double standard


Observing the Web as a Medium and Mirror of (anti-)social Practices

• (Self-)Expression

• Dark Web– Crime– Gold farming– Violence– Pornography– Identity theft

• Sex lifes– Fetishes– prostitution

• Relationships– Breakups– Mobbing– Stalking– Advising– Counseling– Democracy


The Web as Medium & Mirror of NEW (?) (anti-)Social Practices

Automation• Twitter bots• Online dating bots• Smart contracts

(„Code is law“)– E.g. DRM

• Quantified everything– Physique – Psyche

• Infer your Big 5 personality traits from your facebook profile

• Threats to Internet Security & Safety– Hacking – Social engineering


Produce

Consume

Cognition

EmotionBehavior

SocialisationKnowledge

Observable Micro-

interactions in the Web

AppsProtocols

Data & InformationGovernance

WWW

Observable Macro-

effects in the Web

Web Science: Discipline or Transdisciplinary Endeavour?


How to investigate the Web?


Web Observatories Konect


Why to observe?• Understanding

– Collecting– Describing– Analyzing– Modeling– Predicting– Repeating!


Challenges – Data Collection IssuesLegal and/or Ethical • Crawling

– May be disallowed by provider

• Usage logging– Privacy of individuals

• Even if it is allowed....


Challenges – Data Collection Issues• Crawling

– What does it mean to crawl a heavily interactive site?– Incomplete data

• Unreachability• Time outs



– What does it mean to crawl a heavily interactive site?– Incomplete data– Where to start?

• We cannot observe everything!– Even just for data size!– What appear to be most fruitful starting points?



– What does it mean to crawl a heavily interactive site?– Incomplete data– Where to start?– Where to stop?

• Each crawl is a view– Twitter

» Tweet• URL

• Web Page• Subweb

» Followers• Followers‘ Followers

• ...



– What does it mean to crawl a heavily interactive site?– Incomplete data– Where to start?– Where to stop?– Synchronous vs asynchronous

• Strictly speaking: only asynchronous crawling possible– But in [Dellschaft&Staab] we targeted the construction of models for

streams of tags


Challenges – Data Publishing IssuesLegal and/or Ethical Example Issues• AOL query log• Netflix challenge• Delicious

– http://www.tagora-project.eu/data/• Twitter

– Collecting, but no sharing• SocialSensor project

http://www.tagora-project.eu/data/

http://www.tagora-project.eu/data/


Challenges – Data Publishing IssuesTechnical/Modelling issues• Generic format, e.g. RDF• Format ready for digestion by a certain software, e.g.

for Matlab processing• Openness to other data

– E.g. references to DBPedia/Wikipedia• Accuracy of publishing

– http://me.org showed „...“– http://me.org showed „...“@2013-05-01:0900CEST– http://me.org showed „...“@2013-05-01:0900CEST called

from IP 193.99.144.85 using browser...version...history...

http://me.org/

http://me.org/

http://me.org/


Sharing Software• Software

– For crawling or usage logging– Rather than sharing the data, share the code for observing

• Example: code for crawling Twitter in a certain way

• Issues– Limited repeatability– Disturbance liability („Störerhaftung“) – at least in DE

• If you provide source code for crawling, e.g., Facebook, even if you do not crawl FB, FB can sue you


More later by Jerome


Model the Web

What is in the content?

What is in the algorithm?

What is in the Social machine?

Web

Obs

erva

tory


Example Topic: Bias


Bias in the Device

Accessibility• Impaired eyesight• Impaired use of

mouse and keyboard• ...

HTML5 semantics helps – but is not used much


Bias in the Device

Accessibility• Impaired eyesight• Impaired use of

mouse and keyboard• ...


Haris Aslanidis


Bias in the Device

Accessibility• Impaired eyesight• Impaired use of mouse and

keyboard• ...


http://west.uni-koblenz.de/en/research/projects/mamem

http://www.mamem.eu/mamem-makes-available-gazetheweb-browse-application/










Search engines• Categorizing people and animals

– White vs black– http://www.nytimes.com/2016/06/26/opinion/sunday/

artificial-intelligences-white-guy-problem.html?_r=0

• Job advertisements– Well-paid job not offered to females

Bias in the Software


Bias in Content/Data

Credit Hire Sex Ethnic Zip Height ... ...

+ +

+ -

- +

+ +

- -

correlated

Data protection laws suggest not to process sensitive data attributes

like „sex“ or „ethnic“


Example:

Notable women described by „has husband“

Notable men not described by „has wife“


Bias in Content: Social Networks(Lerman et al 15)


fish, rice

seafood, fish seafood, shrimp lobster, wine

seafood, fish, salmon

fish, salmon, wine

rice, fish

lobster, seafood, shrimp

coffee

coffee, wine

coffee

wine

wine

pizza, wine

pizza, wine

pasta, wine

pasta, shrimplobster, shrimp

seafood, shrimp

Tagged photos with geo-coordinates from Flickr


fish, rice

seafood, fish seafood, shrimp lobster, wine

seafood, fish, salmon

fish, salmon, wine

seafood, shrimp

lobster, seafood, shrimp

coffee

coffee, wine

coffeeitalian, wine

wine

pizza, wine

italian, pizza, wine

pasta, wine

pasta, shrimp

seafoodfishlobstershrimpcrabwinesalmon

winepizzacoffeeitalianpasta

seafood, shrimp

lobster, shrimp

Bias in the Algorithm: Shape of Clusters


Evaluation: Anectodal, Perplexity, Gaming

Gaming study: intrusion detection

Precision 8 topicsavg / median

LGTA 0.60 / 0.58

Basic model 0.64 / 0.58

MGTM 0.78 / 0.75


• The Web reflects current and past discrimination!

• Example:– Predict who will leave the company based on email

features (ICWSM-16)• People with outsiders‘ vocabulary more prone to fail• Unresolved:

– Do they fail because they are less adaptable or– Do they fail because the environment is hostile to outsider?

Bias and Social Practices


Wikipedia• Efforts to counter bias

Law• E.g. UK equality act• Protected characteristics:

– Age, disability, gender, marriage, religion,...

Bias and Processes


Biases in the Social Machine:The Case of Liquid Feedback


...


Online Delegative Democracy

CC-BY-SA Ilmari Karonen


Delegative Democracy

• Between direct and representative democracy





• Voters can delegate their vote to other voters







• Delegations can be revoked at any time






• Delegations can be revoked at any time

• Votes are public!


Dataset:LiquidFeedback

(German Pirate Party)


LiquidFeedback – Pirate Party• Observation: 08/2010 – 11/2013

• 13,836 Members• 14,964 Delegations

• 499,009 Votes


LiquidFeedback – German Pirate Party

•

Users create initiatives, which are grouped by issues and belong to areas



•


Area: Environmental issuesIssue: CO2 output has to be reduced.Initiative: Subsidise wind turbines!



•


Area: Environmental issuesIssue: CO2 output has to be reduced.Initiative: Subsidise wind turbines!

Areas: 22Issues: 3,565Initiatives: 6,517



• Users create initiatives, which are grouped by issues and belong to areas

Delegations on global, initiative, issue and area level

→ “Back-delegations” possible


Dataset – First Impressions

•


Dataset – First Impressions

•

Voting Weight


Dataset – Bias ?

•

3,658 members > 10 votes1,156 members > 100 votes 54 members > 1,000 votesMedian all: 8 votes

Median delegating: 42 votesMedian delegates: 64 votes


Delegation Network• Temporal analysis

•



•


Investigation• What are the social processes?• How are data collected?• Which algorithms are used?• How is the algorithm used in system processes?• What are the system effects?

Purposes of investigation• Social sciences: Observe bias in action!• Intervention: Change the overall system!• System sciences: Understand system limits!• Engineering: Build „proper“ technical system!

Summary: Bias


Web Models


• Descriptive– Qualitative– Statistical

• Predictive– Modeling deterministic regularities

• Generative– Modeling non-deterministic principles

• Liking a song• Creating a link

Web Models


Descriptive ModelsExample: Bow Tie Structure of the Web


Bow-tie structure of the Web


Predictive ModelsExample: Link Prediction by Triangle Closing


Social Network

Person Friendship


Recommender Systems

Predict who I will add as friend next

Standard algorithm: find friends-of-friends

me


Friend of a Friend

1 2 4 5 6

3

Count the number of ways a person can be found as the friend of a friend.


Generative ModelsExample: Link Creation by Barabasi-Albert


• Many large networks are scale free– Matthew effect: rich get richer

• Vote delegation!

• The degree distribution has a power-law behavior for large k (far from a Poisson distribution)

• Random graph theory and the Watts-Strogatz model cannot reproduce this feature

General considerations


Scale free

Same shape of hull, no matter which resolution


Scale-free of Web distribution

Same shape of distribution: no matter which k


Two generic mechanisms common in many real networks:• Growth (www, research literature, ...)• Preferential attachment:

attractiveness of popularity

The two are necessary

Barabasi-Albert model (1999)

Barabási & Albert, Science 286, 509 (1999)


• t=0, m0 nodes• Each time step we add a new node with m (m0)

edges that link the new node to m different nodes already present in the system

Growth


• When choosing the nodes to which the new connects, the probability that a new node will be connected to node i depends on the degree ki of node

Preferential attachment

( ) ii

jj

kk

k

Linear attachment (more general models)Sum over all existing nodes


Numerical simulations

• Power-law P(k)k- SF=3• The exponent does not depend on m (the

only parameter of the model)


The Past and the Future Web


• 1945 Vannevar Bush, „As we may think“, Memex• 1962 Ted Nelson, Hypertext• 1965 Wide area network• 1968 Doug Engelbart, The mother of all demos• 1972 Public Arpanet, Email• 1974-82 Internet protocol TCP/IP• 1978 Consumer information services & Email• 1983 AOL, online service for games, communities...• 1984 Domain name service

Pre-Web


The World Wide Web• 1989 Concept drafted by Tim Berners-Lee• 1993 National Center for SuperComputing

Applications launched Mosaic X• 1994 First WWW conference• 1994 W3C started at MIT• Commercial websites began their proliferation• Followed by local school/club/family sites• The web exploded

– 1994 – 3,2 million hosts and 3,000 websites– 1995 – 6,4 million hosts and 25,000 websites– 1997 – 19,5 million hosts and 1,2 million websites– January 2001 – 110 million hosts and 30 million websites


The World Wide Web– 1994/1995 Amazon– 1994/1995 Wiki– 1995 AltaVista Search Engine– 1995 Internet Explorer– 1997-2001 Browser wars– 1996-1998 XML recommendation– 1998 Google– 1999 First W3C recommendation on RDF (Semantic Web)– 2001 Dot.Com bubble bursts– 2001 Wikipedia– 2003/2004 Facebook– 2004 Flickr– 2005 YouTube


Concepts Example Applications

Web of People Physical transport service (Uber (2009), Lyft), accommodation service (AirBnB, Couchsurfing), online dating service,

2009

Web of Things Smart city, ambient intelligence, personal and public health information, personal and public transport information

2007

Web of Services Cloud services, Digital transformation, Programmable Web (2005)

2005

Web of Data User generated content applications (Facebook, Wikipedia (2001) and Wikidata,…), Linked open gov data

2001

Web of Documents HTTP, HTML, XML, Browser (Mosaic 1993)

1993

Computer Networks/Internet

Document delivery (internet 1982), VOIP, Streaming

1982


Internet of Things vs Web of Things

IoT• Internet• P2P networks• Sensors

Web of things• How to link?• What is „linking WoT“?• How to use by me?• How to discover?• How to search?


Web of People


http://static.tapastic.com/cartoons/19/f5/11/a0/a70ea5d7dbdd477d9e3bc2e0a2bfa286.gif


Aaron Swartz †

Co-author of RSS1.0 at the age of 14


http://static.tapastic.com/cartoons/19/f5/11/a0/a70ea5d7dbdd477d9e3bc2e0a2bfa286.gif


Web of People

• What is it?– Identification

• Orcid• Facebook ID• Oauth 2.0?

– Trust• Uber score• Airbnb score• Ebay score• Tinder score• Group score

– Chinese social score– Credit rating score

?


Concepts Delivered technical capabilitiesWeb of People Identification and rating by/of peopleWeb of Things Identification, linking, aggregation,

monitoring and controlling of thingsWeb of Services Identification, composition and calling of

servicesWeb of Data Identification, linking and retrieval of dataWeb of Documents Identification, linking and retrieval of

documentsComputer Networks/Internet

Identification of and communication between computers


Concepts StandardsWeb of People No mature standards yetWeb of Things No mature standards yetWeb of Services REST, JSON, JSONLDWeb of Data RDF, SPARQLWeb of Documents HTTP, HTML, XML, AJAXComputer networks/Internet Internet, TCP/IP, Optical fibre,

5G


Your thoughts?


Conclusions


What is in the data?

What is in the algorithm?

What is in the Social machine?

Web

Obs

erva

tory


Telling the Story with Web Science

What is in the Data?

What is in the Algorithm?

What is in the Social Machine?

Story telling

Under-standing

Modelling


Web

Accomplishments

• Web of Services• Web of Data• Web of Documents• Computer

networks/Internet

Future in the making

• Web of People• Web of Things

How to identify and use?How to observe?How to resolve issues?


Thank you!