Over the past decade, computers and networks have become increasingly ubiquitous. They have a central role in how we work, communicate, and learn. In this talk, I introduce the concept of human-machine collaboration and show how networked information systems enable coupled relationships between humans and machines. I will take four perspectives on this coupled relationships: how we can design, analyze, develop and also extend them. The presented approaches contribute to the research field of networked socio-technical information systems. By unfolding design parameters of these systems, we can develop adaptive networked information systems that, in the future, can reduce the friction between humans and machines to a point where they become a natural extension of our human experience.

1. Claudia Müller-Birn
Institute of Computer Science, Freie Universität Berlin
March 2, 2017
Human-Machine Collaboration
in Networked Information Systems

2. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 2
Social Web
that connects
people and knowledge
Semantic Web
that connects
machines and knowledge
…
…
Human-Machine
Collaboration

3. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 3
Research Context
Networked information systems are socio-technical systems
that integrate distributed information resources
based on usable user and programming interfaces
to enable human-machine collaborations.

4. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Research Areas
We focus on designing networked information systems for both, human and
algorithmic agents in the following three areas:
(1) Designing efficient, effective and easy-to-use interfaces.
(2) Enabling a seamless coordination between knowledge processes.
(3) Expanding collaboration into the physical world.
4

5. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Work Location 1
5
The "almost" complete HCC:Team (Summer 2016)

6. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Work Location 2
6

7. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 7
Working Process
Review and Analysis
Ideas and Concept
User Scenario
DesignPrototyping
Testing
Implementing

8. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Designingnetworked information systems
Analysingrelationships between humans and machines in networked
information systems
Developingcoupled relationships between humans and machines in
networked information systems
Extending networked information systems
8

9. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Designingnetworked information systems
Analysingrelationships between humans and machines in networked
information systems
Developingcoupled relationships between humans and machines in
networked information systems
Extending networked information systems
9

10. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 10
Challenges in online communities
Everyone is better off
if everyone contributes
than if no one does
Each individual is even better off
if she does not contribute
while the others do
Critical mass
Social loafing

11. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Feedback: Goal setting
Context: MovieLens is a movie recommendation site
(http://www.movielens.org)
Hypothesis: In an online community, specific, numeric
goals will motivate greater contributions than
non-specific goals
Approach: Members were recruited by sending them an
email message which contained an invitation to participate
in a movie rating campaign
Result: Providing community members with specific goals increases their contributions.
11
Ling, Kimberly, et al. "Using social psychology to motivate contributions to online communities." Journal of Computer-Mediated Communication 10.4 (2005).

12. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Sociality: Social contact
Context: Wikipedia
Hypothesis: Social contact improves the
retention rate of new editors
Approach: Create a online teahouse for
meeting new editors
Result: Social contact with other similar
contributors causes members to stay and to
contribute more.
12
Wikimedia, 2012. http://en.wikipedia.org/wiki/Wikipedia:Teahouse

13. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Feedback: Make readers’ interests visible
Constant interest
Regularly accessed articles, sometimes only for fact
finding, e.g. Albert Einstein, Facebook, IMDB
13
Peak interest
Death of people, game and movie releases, e.g. Whitney
Houston, The Hunger Games, 2012 Phenomeno
Increasing/decreasing interest
Items that became popular/loose popularity during our
observation period, e.g. One direction, Instagram
Lehmann, J.; Müller-Birn, C.; Laniado, D.; Lalmas, M.; Kaltenbrunner, A. (2014): Readers Preference and Behavior on Wikipedia. In Proceedings of the 25th ACM Conference on Hypertext and
Hypermedia, Santiago (Chile).

14. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Designingnetworked information systems
Analysingrelationships between humans and machines in networked
information systems
Developingcoupled relationships between humans and machines in
networked information systems
Extending networked information systems
14

15. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Research Goal
Funding
Data Collection
Analysis of Contribution Patterns
» Identifying participation patterns for collaborative ontology development
» DFG-funded research project in collaboration with GESIS - Leibniz-Institute for
Social Science
» JSON and XML-based data dump (10/2012 to 10/2014)
15
Wikidatameter by Lukas Benedix (2013)

16. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 16
“What is the population of Berlin?”

17. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 17
project launch in 2012 | over 460 million edits | 25 million items

18. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 18
Human Contribution Patterns
Created items
Item terms
Item statements
Set sitelinks
Created properties
Property terms
Discussions
Protected pages
Reverts
Müller-Birn, C., Karran, B., Lehmann, J. and Luczak-Rösch, M., 2015. Peer-production system or collaborative ontology engineering effort: What is Wikidata?. In Proceedings of the 11th
International Symposium on Open Collaboration (p. 20). ACM.

19. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 19
Contributions on German Wikipedia
Müller-Birn, C., Dobusch, L. and Herbsleb, J.D., 2013, June. Work-to-rule: the emergence of algorithmic governance in Wikipedia. In Proceedings of the 6th
International Conference on Communities and Technologies (pp. 80-89). ACM.

20. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 20
Human and Algorithmic Contribution Patterns
Created items
Item terms
Item statements
Set sitelinks
Created properties
Property terms
Discussions
Protected pages
Reverts
?
Müller-Birn, C., Karran, B., Lehmann, J. and Luczak-Rösch, M., 2015. Peer-production system or collaborative ontology engineering effort: What is Wikidata?. In Proceedings of the 11th
International Symposium on Open Collaboration. ACM.

21. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 21
Müller-Birn, C.; Dobusch, L.; Herbsleb, J.D. (2013): Work-to-rule: the emergence of algorithmic governance in Wikipedia. In Proceedings of the 6th
International Conference on Communities and Technologies (C&T '13). ACM, New York, NY, USA, 80-89.
0
1,000,000
2,000,000
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
boteditcountperyear
Category
Main
Other
Portal
Talk
User
User talk
Wikipedia
editing activities includes
additional namespaces
editing activities extends to
community space
focus expands further to
project spaces about 45 % of all
edits are outside
of article
namespace

22. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 22
Algorithmically translated rules are more scalable, but
they are less likely to handle exceptions.
Müller-Birn, C.; Dobusch, L.; Herbsleb, J.D. (2013): Work-to-rule: the emergence of algorithmic governance in Wikipedia. In Proceedings of the 6th
International Conference on Communities and Technologies (C&T '13). ACM, New York, NY, USA, 80-89.
Changing task focus of a bot over time

23. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 23
Algorithmic Governance - how bots influence user retention
Halfaker, A., Geiger, R.S., Morgan, J.T. and Riedl, J., 2012. The rise and decline of an open collaboration system: How Wikipedia’s reaction to popularity is causing its decline. American Behavioral
Scientist, p.0002764212469365.

24. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Designingnetworked information systems
Analysingrelationships between humans and machines in networked
information systems
Developingcoupled relationships between humans and machines in
networked information systems
Extending networked information systems
24

25. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Research Context
Research Goal
Context
Integrating Interdisciplinary Research Data
» Research Data in the area of biological research
» Allow experts to discover research data beyond disciplinary
terminologies
» Research project with 20 partners; in close collaboration with Botanical
Garden and Botanical Museum (BGBM)
25
Karam,N.;Müller-Birn,C.,Gleisberg,M,;Fichtmüller,D.;Tolksdorf,R.Güntsch, A. (2016): A Terminology Service supporting semantic annotation, integration, discovery and analysis of
interdisciplinary research data. Datenbank-Spektrum. Zeitschrift für Datenbanktechnologien und Information Retrieval, Vol. 16, No.3, 195-205.

26. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 26
German Federation for Biological Data (GFBio)
Karam,N.;Müller-Birn,C.,Gleisberg,M,;Fichtmüller,D.;Tolksdorf,R.Güntsch, A. (2016): A Terminology Service supporting semantic annotation, integration, discovery and analysis of
interdisciplinary research data. Datenbank-Spektrum. Zeitschrift für Datenbanktechnologien und Information Retrieval, Vol. 16, No.3, 195-205.

27. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 27
Authmann, C., Beilschmidt, C., Drönner, J., Mattig, M. and Seeger, B., 2015. VAT: a system for visualizing, analyzing and transforming spatial data in science.
Datenbank-Spektrum, 15(3), pp.175-184.

28. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 28
The Terminology Service
CHEBIBCO
PATO RECORDBASIS
ENVO
OBOE
KINGDOMQUDT
SWEET
ISOCOUNTRIES
LIT_I NCBITAXON
PESI
COL
ITIS
…
External
Terminologies
Karam,N.;Müller-Birn,C.,Gleisberg,M,;Fichtmüller,D.;Tolksdorf,R.Güntsch, A. (2016): A Terminology Service supporting semantic annotation, integration, discovery and analysis of
interdisciplinary research data. Datenbank-Spektrum. Zeitschrift für Datenbanktechnologien und Information Retrieval, Vol. 16, No.3, 195-205.

29. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 29
Terminology Matching
CHEBIBCO
PATO RECORDBASIS
ENVO
OBOE
KINGDOMQUDT
SWEET
ISOCOUNTRIES
LIT_I NCBITAXON
Abbreviation Name
#
classes
# same
URI’s
NCBITAXON
National Center for Biotechnology Information (NCBI)
Organismal Classification
1.507.562
CHEBI Chemical Entities of Biological Interest Ontology 103.755
ENVO Environment Ontology 6.190
BCO: 11
PATO: 221
SWEET
Semantic Web for Earth and Environment Technology
Ontology
4.549
ISOCOUNTRIES ISO 3166 Countries and Subdivisions 5.307
PATO Phenotypic Quality Ontology 2.603 ENVO: 221
LIT_I The lithologs rock names ontology for igneous rocks 1.870
OBOE Extensible Observation Ontology 538
QUDT Quantity, Unit, Dimension and Type 204
BCO Biological Collections Ontology 127 ENVO: 11
RECORDBASIS GFBio Agreed Vocabulary for RecordBasis 14
KINGDOM GFBio Agreed Vocabulary for Kingdoms 9
PESI
COL
ITIS
…
External
Terminologies

30. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Designingnetworked informaton systems
Analysingrelationships between humans and machines in networked
informaton systems
Developingcoupled relationships between humans and machines in
networked information systems
Extending networked information systems
30

31. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Research Context
Research Goal
Context
Collaborative, geographically distributed research processes
» Many collaborative research practices evolve around physical artefacts,
e.g. in archeology, and are geographically distributed
» Integrating physical and digital activities more smoothly and allow for
computer-supported collaborative sensemaking by
networked information systems
» Research project Hybrid Knowledge Interaction of the Cluster of Excellence
Image, Knowledge Gestaltung. An Interdisciplinary Laboratory
31

32. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017 32
Computer-Supported Collaborative Sensemaking
Software Agents
Cognitive Process of Person 2
Representation Shift Loop
Search
for good representations
Instantiate
Representations
Cognitive Process of Person 1
Representation
Interface
Processing Requirements of Task
Generation
Loop
Data
Coverage
Loop
Representation Shift Loop
Search
for good representations
Instantiate
Representations
Generation
Loop
Data
Coverage
Loop
Processing Requirements of Task
Set of Annotations
External Knowledge Bases
Hong, M. and Müller-Birn, C., 2017, February. Conceptualization of Computer-Supported Collaborative Sensemaking. In Proceedings of the 20th ACM Conference Companion on Computer
Supported Cooperative Work & Social Computing. ACM.

33. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Conclusion & Future Work
» The presented qualitative knowledge-based workflows build upon Licklider's vision of a “man-
machine symbiosis” and Engelbart’s “augmenting human intellect” proposal
» The technical and human components of a web information system are equally important,
whereby the design of the interfaces and the human ability to use them should be coordinated.
» The algorithmic component should be implemented in a way that it allows users to get an
understanding of their functioning
33

34. Prof. Dr. Claudia Müller-Birn | Human-Machine Collaboration | March 2017
Discussion
34