To cut expenses and save time, enterprise crowdsourcing is more and more used to disseminate corporate tasks, which are traditionally performed by a small group of people, to an undefined large workforce within and beyond the boundaries of a company. However, harnessing the positive effects of crowdsourcing faces several challenges, such as the efficient and proper assignment of a crowdsourcing task to an available and competent group of workers, or the securing of the integration and reuse of crowdsourcing data across heterogeneous business applications. To overcome these challenges, a semantic standard for enterprise crowdsourcing is developed and its applicability is shown by evaluating it against three diverse scenarios that may occur in real business environments. The proposed standard finally includes fifteen semantic elements to describe a crowdsourcing task and eight elements to define a user.
Towards a semantic standard for enterprise crowdsourcing
1. Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
ECIS, Utrecht, June the 7th, 2013 Lars Hetmank
Towards a Semantic Standard for Enterprise
Crowdsourcing
A scenario-based evaluation of a conceptual
prototype
2. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 2 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Short Vita
Lars Hetmank
PhD student at Dresden University of
Technology
Chair of Business Informatics, especially
Information Management
June 7th, 2013
Research interests:
• Information and Knowledge Management
• Semantic Web and Linked Data
• Crowdsourcing
4. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 4 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
What are the Challenges for Enterprise
Crowdsourcing?
June 7th, 2013
EXTERNAL BUSINESS APPLICATIONS
CROWDSOURCING SYSTEM
Crowdsourcing Task CrowdRequester
12 Task allocation
11 Task specification
propose
select
define
13 Team formation
14 Monitoring adjust process
Knowledge
repository
Enterprise
SNS
HR
database
15 Interoperability
...
6. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 6 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Methodology
June 7th, 2013
Application of design-science research (DSR) guidelines according to Hevner et al., 2004
• Design-science research approach (Hevner et al., 2004)
• Based on a systematic literature review on crowdsourcing
systems (Hetmank, 2013)
Design as
an Artifact
Problem
Relevance
Research
Contribu-
tion
Design
Evaluation
Research
Rigor
Search
Process
Research
Communi-
cation
Semantic
standard
Enterprise
Crowd-
sourcing
Metadata
Schema
Automation
and inter-
operability
First proof-of-
concept
Scenario
building
Based on
previous
studies on
crowds-
ourcing
First step in
the develop-
ment process
Technical-
oriented and
management-
oriented
audience
Conferences,
Journals, and
Prototypes
7. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 7 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Suggestions for Improvements …
① What additional requirements must an
enterprise crowdsourcing metadata schema
meet?
② What classes, relationships, attributes
are missing or may not be required?
Answer live at PollEv.com/ecms !
June 7th, 2013
9. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 9 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
June 7th, 2013
Task description
Target audience
Complexity
Type of action
Modularization
Type of the
reward
Nature of the
reward
Latency Duration
Submission time Closure time
Visibility
Confidentiality
Human
requirement
Technical
resource
PollEv.com/ecms
?
Crowdsourcing Task
10. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 10 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
June 7th, 2013
User identity Nationality
Qualification
Job title
Entry date
Department Location
Accomplishment
PollEv.com/ecms
Crowdsourcing User
12. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 12 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Evaluation and 1st Proof-of-concept
Three scenarios that may occur in real
business environments:
1. Evaluate product design proposals
2. Translate technical specification
3. Build company-wide virtual library
Four users as a subset of an example crowd
June 7th, 2013
PollEv.com/ecms
13. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 13 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Element Scenario 1
Task description Evaluate product design
Target audience Hybrid
Complexity Simple
Type of action Evaluate
Modularization 10 subtasks (bundled)
Latency Immediate
Nature of the reward Fixed and performance-based
Type of the reward 15 reputation points plus
bonus or discount of 5
Submission time After release
Closure time After 20 reviews for each
product design
Duration 1 minute
Confidentiality Low
Visibility Hidden
June 7th, 2013
Element User 1 User 2
User identity Alan Coulter Adèle Girard
Location Cork Lyon
Nationality Irish French
Job title Chief product
designer
Junior product
engineer
Entry date 1993-04-01 2010-02-09
Department Product
development
Product
engineering
Qualification Master of Product
Design and
Development
Bachelor of
Engineering
Accomplish-
ment
http://example-
company.com/
cs/task/3241
<none>
Human requirement Job tenure of more than two
years OR
master in engineering, product
design, marketing OR sales
Technical resource http://www.flickr.com/
photos/new-product-xyz
15. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 15 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Future Improvements and Research
• some elements are currently oversimplified and require
further refinement in their level of detail
• facilitate the definition of conditional expressions
• create additional concepts (contribution, evaluation and
reward mechanism)
• reuse of existing standards, such as
FOAF, schema.org, Dublin
Core, activitystrea.ms, SIOC, Proton, HR XML, WS-BPEL
Extension for People (BPEL4People), or XMP Process
Definition Language
June 7th, 2013
PollEv.com/ecms
16. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 16 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
Conclusion
• foster the standardization in the domain of enterprise
crowdsourcing
• 5 current challenges
• design-science research approach
• 1st conceptual prototype
• crowdsourcing task (15 metadata elements)
• crowdsourcing user (8 metadata elements)
June 7th, 2013
PollEv.com/ecms
18. Towards a Semantic Standard for Enterprise Crowdsourcing Slide 18 | 18
Faculty of Business and Economics | Chair of Business Informatics, especially Information Management
June 7th, 2013
Towards a Semantic Standard for
Enterprise Crowdsourcing
A scenario-based evaluation of a conceptual prototype
Answer live at PollEv.com/ecms !
SKYPE
larshetmank
TWITTER
twitter.com/larshetmank
EMAIL
lars@hetmank.de
MOBILE
+49 176 24 77 39 65
WEBSITE
www.larshetmank.com