Proposing a System to Support Crowdsourcing

13
Proposing a System to Support Crowdsourcing
Manuela Aparicio
Adetti - IUL/ISCTE - IUL
Av. Das Forças Armadas
Lisboa, Portugal
manuela.aparicio@iscte.pt
Carlos J. Costa
Adetti – IUL /ISCTE - IUL
Av. Das Forças Armadas
Lisboa, Portugal
carlos.costa@iscte.pt
Andrew Simoes Braga
University of Winnipeg
Winnipeg, Manitoba,
Canada
braga.a@webmail.uwinnipeg.ca
ABSTRACT
In this paper, a conceptual framework is proposed, supported in
the literature review, derived by identifying the main concepts
related to crowdsourcing, as well as ways of improving group
participation. We also propose a software solution that may be
used to support the crowdsourcing process. This software solution
is inspired by the conceptual framework.
Categories and Subject Descriptors
H.5.3 [Information Interfaces]: Group and Organization
Interfaces, Collaborative computing, Web-based interaction.
General Terms
Management, Economics, Human Factors
Keywords
Crowdsourcing, Collective intelligence, Web Application, Web
Based Systems, Collaborative Systems
1. INTRODUCTION
Our daily lives are filled with interactions as members of a group,
a professional team, or as individuals who make connections and
contribute to society. People use web to communicate, to work
and to have leisure time. People are increasingly using the web to
communicate for work purposes, as well as in their leisure time.
This has increased the amount of time people are spending online
every day, while also increasing its importance with the evolution
of social system networks. Cooperation is spreading around the
globe, linking people from different countries, cultures and
backgrounds. Web applications are nowadays available in various
formats, and it has become quite common for individuals to check
their email or social network accounts on their smart phones and
tablets many times a day. Collaborative websites are being
developed with each individual's free contributions, and
companies are willing to use the crowds as a rational resource to
sovling their problems. In this context we purpose a system that
aims to support crowdsourcing.
In this paper we propose a conceptual framework supported on
the basis of our review of the existing literature. We also present a
solution to support crowdsourcing.
In the following section, we identify the main concepts related to
crowdsourcing, as well as ways of improving group participation.
We then present a generic framework, and finally a specific
solution inspired by this framework.
2. LITERATURE REVIEW
Collective intelligence is “a form of universally distributed
intelligence, constantly enhanced, coordinated in real time, and
resulting in the effective mobilization of skills” [7]. Collective
intelligence is also defined by Howe [5] as a “form of group
cognition” in which each individual performs his own role, taking
part in a larger community of players. Howe compares the
integration of knowledge derived from different people that
communicate within a group in order to achieve certain
objectives. Malone and colleagues [8] created a model for
mapping the genome of collective intelligence, in which it is
defined as “groups of individuals doing things collectively that
seem intelligent”.
James Surowiecki [12] refers to two different ways of
understanding what he calls the wisdom of crowds. Surowiecki
[12] defines the wisdom of crowds as decentralized wisdom
generated by a crowd of self-interested people, working in a
decentralized way on the same problem. Crowdsourcing emerged
from a new paradigm, involving external people in the
organization in order to be creative and to collaborate on solving a
problem, or to pursue an objective [6]. The crowdsourcing
concept is based on the principle that each individual has
knowledge that is valuable to other people.
In literature (e.g. [6], [10]) some crowdsourcing types arise:
 Prediction market or information market
 Problem solving or crowdcasting
 “Idea Jam”
The prediction, or information market relies on public opinion,
and therefore pools from “futures” purchases to make decisions in
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies
are not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
OSDOC'12, June 11, 2012, Lisbon, Portugal.
Copyright 2012 ACM 978-1-4503-1284-4/12/0006..$10.00.

14
different situations. Problem solving, or crowdcasting, is a type of
crowdsourcing that some organizations use to get inputs to the
problems that arise in their everyday operation. The “Idea Jam” is
also used, mainly by organizations when faced with the need to
get new and fresh ideas using crowd brainstorm suggestions.
Crowdsourcing models are connected to the referred typology [6].
Crowdcreation is a model used to channel the energy of a crowd
in an outsourced way, like when companies use the crowds for TV
advertisements, language translation, or any other form of co-
creation, by adding individual parts to build a whole. The Crowd
voting model can be applied to various situations, such as when
organizations need judgment on ideas, for public opinion on
controversial political matters, or even when search engines
organize outputs according popularity. Another model is
crowdfunding, which is a way that public organizes in order to
raise funds for a common cause.
In order to have a crowd be part of a collaborative community, it
is mandatory to recruit users, and to retain them [3]. Existing
literature points out some useful strategies[3]:
 Require users
 Pay Users
 Ask for volunteers
 Make users pay for a service
 Piggyback on the user traces
There are five ways user recruiting can be identified. First, we can
require users if we have the authority to do so. Second, users can
be paid to perform specific tasks. Third, we can apply for
volunteers, which is, in fact, the most popular method for many
successful websites like Wikipedia, Linux, Yahoo, and You Tube,
among others. Fourth, users can be asked to pay for a service. For
example, users can leave a comment on a system only after they
solve a puzzle. The puzzle/Captcha solving can be the payment
from the user to the system by reCaptching. When the user retypes
a word that an OCR program has failed to recognize in system A,
the retyping is a “payment” that consists in an effort to digitize
written text into system B. The fifth way to recruit users can be
accomplished by tracing users on a specific system. For example,
building a spelling correction system by exploiting user traces of a
search engine, like Google or Yahoo!
Once users are part of a crowd system, they should feel
encouraged to stay and be part; In other words, systems should
retain users. Doan and colleagues [3] identified the main popular
methods of user retention:
1. Provide instant graphics
2. Provide an enjoyable experience or a necessary service
3. Establish a measure to show fame/trust/reputation
4. Set up a competition
5. Provide an ownership situation
On one hand, systems can provide instant gratification by
showing to the user how their participation was valuable. Another
way of retaining users is to provide an enjoyable experience while
they make their contribution. Yet Another is to enable users to
measure of their fame, popularity, or reputation in the community,
or to set competitions between users, showing top rated users.
Ownership can also be a solution to cultivate the feeling in the
user that they belong to, or own, part of a system.
On the other hand, the contexts in which this concept can be used
are very wide, including art creation [13], [14]. Smart phones, by
integrating the Web with mobile phones, have become essential -
not only for communicating, but also for sharing implicitly and
explicitly generated information [15] [16]. Those opportunities
also present new challenges, for example, new forms of
presenting information and supporting crowdsourcing processes
[16].
3. A FRAMEWORK
In order to analyze the main issues related to crowdsourcing, and
to create new paths of development, we propose here a conceptual
framework. In this framework we identify the participants (who),
the tasks (what), motivations and incentives (why) and the ways in
which tasks are accomplished (how).
Table 1 – A Framework to Identify
Who What Why How
Problem
proponents
Problem Incentive to
propose a
problem
Anonymity
Access to
Information
Idea
Generator
Solution Incentive to
propose a
solution
Anonymity
Access to
Information
Evaluators Vote Incentive to
vote
Anonymity
Access to
Information
Who- Who are the main participants in the process? The
participants may be problem proposers, idea proposers and/or
evaluators. Each one of those roles may be played by participants
with specific skills.
What- As already mentioned (e.g. [3]), there are several forms of
participation. McGrath [9] built on Hackman's [4] observations,
describing four general processes: Generating, Choosing,
Negotiating, and Executing. Specifically, in this context we
identified three tasks: proposing a problem, proposing a solution
and voting. The first two tasks correspond to generating, in
McGrath's processes, while voting corresponds to the Choosing
process.
Why- A person's motivation to contribute is one of the most
important issues related to crowdsourcing. In fact, as Howe [6]
points out, getting people involved requires understanding what
motivates them to contribute. He also stresses that people need to
feel rewarded for their efforts, even if it takes the form of a small
amount of money. On the other hand, the incentives also depend
on the entity that is requesting participation [6].
How- There are several ways of participating. They depending not
only on the task (e.g. generating, choosing), but it also change
according to the rules that participants agree to follow. The degree
of participant anonymity, the level of information sharing, and the
type of incentives involved may change how people participate.

15
4. IMPLEMENTINGA SOLUTION
The proposed system supports different types of users, with
different motivations for contributing to a problem situation/
scenario. The following figure is a use case of the proposed
system.
A user/team/organization may want feedback from multiple
people on a problematic situation. In this scenario, a user has a
proposer profile, and user validated access to the platform, where
they can publish a problem, sharing it with a community.
Someone, or a number of registered users, motivated by the
problem, or by public recognition, or for any other reason,
collaborate, providing their own respective insights into the
problem.
In this system these types of users are called the solvers.
According to their specific configurations, their responses may or
may not be seen by other solvers. Also depending on these
configurations, solvers can see each others' solutions, and
therefore construct their ideas based on them.
The evaluator, still another kind of user, can evaluate the private
and public/shared solutions to the given problem. The evaluator
can be the proposer themselves, but can also be any other public
user who wishes to give their evaluations of the proposed
solutions. In this system, crowdsourcing could be applied to all of
these user types in order that the idea can be supported or
critiqued, or the crowd could vote on ideas as in a wiki concept
[2], [1].
Figure 1- Solution Use Case
In the following class diagram, the five main classes in the system
are identified: User, Solution, Vote, Problem and Comment.
The user may present a problem or a solution. The user may also
vote. As previously noted, the user may be somebody that
presents a problem, comes up with a solution, or evaluates given
solutions.
Users may or may not be anonymous. Participants can remain
anonymous to other participants, but may find it useful to keep
personal information about who participates in order to give
specific rewards.
As presented in the use case diagram (figure 1), the proposer can
customize the system in order that it allows different forms of
public participation. For example, the solution proposed may be
viewed by all the participants or may be accessible only to the
problem proposer.
On the other hand, an important issue is the incentive that the
proposer is willing to give to the participants, according to the
level or quality of participation, or for simply participating. This
issue is very important and subject to future study.
Figure 2- Solution Class Diagram
In previous projects we used JAVA technology, however, due to
the costs involved in maintenance and development, we decided
to use a CMS (Content Management System). This solution
corresponds to the adaptation of CMS into a Crowd Source
System. Specifically, this solution is supported in the Wordpress

16
Table 2 – CroudSourcing System vs. CMS
CrowdSourcing Content Management System
Problem Post
Solution Comment
Vote Rate
Figure 3 is a screen-shot of a publicly shared problem. The figure
specifically shows a list of problems presented in January 2012.
The list is composed of two problems: “How to reduce personal
expenses?” and “Solutions to European crisis”. Those problems
may also be organized by categories. Each user may link several
categories to each problem. Those categories may be useful to
organize the type of problems. In the case we have “social
problems” and “corporate problems”.
Several keywords may also be connected to each problem. For
example, a keyword may be connected to some kind of reward for
users who can solve the problems associated with this keyword.
Figure 3- Inserting Problem
A problem is posted and it is then expected that the public give
feedback by commenting on the problem. The comments will not
be approved (or shared to other participants) to avoid plagiarism.
In figure 4, a user posted a solution to the need of “improving
productivity”.
Figure 4- Inserting Solution
As mentioned before, there are also some users that are
evaluators.
The voting process can also be organized by the proposers of the
problems, in order that their purposes can be achieved.
Figure 5- Evaluating Solutions

17
Development of new modules for Drupal and add-ons for
Wordpress should also be considered.
5. PRELIMINARY EVALATION
This system is currently being used and tested and is in the
development and customization process. Nevertheless, the
problems presented here were created for the purpose of
experimentation.
The use of the system allowed us to improve on some minor
issues related to usability and user experience, however the main
concepts were not seriously scrutinized.
6. CONCLUSIONS
In this paper, we identified the main concepts related to
crowdsourcing, as well as ways of improving group participation.
Supported by the literature, we proposed a conceptual framework
that could be used to either study observed reality, or to support
the development of new systems with new features.
7. ACKNOWLEDGMENTS
This study was partially supported by FCT.
8. REFERENCES
[1] Costa, C. J. (2011). Medieval guild as metaphor to a
knowledge sharing community. Proceedings of the 2011
Workshop on Open Source and Design of Communication,
OSDOC ’11 . New York, NY, USA: ACM. pp 61–64
[2] Costa, C. J., Nhampossa, J. L., & Aparício, M. (2008). Wiki
content evaluation framework. Proceedings of the 26th
annual ACM international conference on Design of
communication, SIGDOC ’08. New York, NY, USA: ACM.
pp 169–174
[3] Doan, A., Ramarkrishnan, R. & Halevy, A. (2011)
“Crowdsourcing Systems on the World-Wide Web” in
Communications of the ACM, April 2011, vol.54. No.4, pp
86- 96
[4] Hackman, J. R., Effects of Task Characteristics on Group
Products. Journal of ExperimentalSocial Psychology (4),
1968, pp. 162-187.
[5] Howe, J. (2006) “The Rise of Crowdsourcing”, Wired
Magazine (14) 6
[6] Howe, J. (2008). Crowdsourcing: Why the Power of the
Crowd Is Driving the Future of Business (unedited edition.).
Crown Business.
[7] Levy, P. (1999). Collective Intelligence: Mankind’s
Emerging World in Cyberspace. (R. Bononno, Trans.). Basic
Books.
[8] Malone, T., Laubacher, R. & Dellarocas, C. (2009)
“Harnessing Crowds:Mapping the Genome of Collective
Intelligence”, MIT Sloan Research Paper 4732-09
[9] McGrath, J.E., "Groups: Interaction and Performance",
Prentice-Hall, Inc, Englewood Cliffs,N.J., 1984
[10] Sloane, P. (Ed.). (2011). A Guide to Open Innovation and
Crowdsourcing: Advice From Leading Experts. Kogan Page.
[11] Sloane, P. (Ed.). (2011). A Guide to Open Innovation and
Crowdsourcing: Advice From Leading Experts. Kogan Page.
[12] Surowiecki, J. (2004). The Wisdom of Crowds. Anchor.
[13] Duarte, P., Costa, C. & Costa, P. (2008) Webstorm: mixing
brainstorming in the web to produce art. Proceedings of the
26th annual ACM international conference on Design of
communication, SIGDOC ’08. New York, NY, USA:
ACM.pp 267-268
[14] Costa, P., Duarte, P. & Costa, C. (2007) WebStorm: mixing
brainstorming with art in the web. Proceedings of the 26th
annual ACM international conference on Design of
communication, SIGDOC ’07. New York, NY, USA: ACM.
pp 170-175
[15] Alt, F., Shirazi, A. ,Schmidt,A. Kramer, U. & Nawaz. Z.
(2010) Location-based crowdsourcing: extending
crowdsourcing to the real world. In Proceedings of the 6th
Nordic Conference on Human-Computer Interaction:
Extending Boundaries (NordiCHI '10). ACM, New York,
NY, USA, 13-22.
[16] Väätäjä, H, Vainio, T.,Sirkkunen,E. & Salo, K (2011)
Crowdsourced news reporting: supporting news content
creation with mobile phones. In Proceedings of the 13th
International Conference on Human Computer Interaction
with Mobile Devices and Services (MobileHCI '11). ACM,
New York, NY, USA, 435-444.

18
The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.

Proposing a System to Support Crowdsourcing

More Related Content

What's hot (11)

Viewers also liked (20)

Similar to Proposing a System to Support Crowdsourcing (20)

More from Carlos J. Costa (20)

Recently uploaded (20)

Proposing a System to Support Crowdsourcing