SlideShare a Scribd company logo
1 of 44
Download to read offline
mCrowd: A framework for incentive based mobile
crowdsourcing
Arpit Jain
Thesis supervisor: Dr. T.V Prabhakar
Department of Computer Science & Engineering
IIT Kanpur
July 13, 2015
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Thesis in a slide
Reasoning
Current incentive based crowdsourcing platforms are not customized
for smartphone users
Most “tasks” on such platforms are lengthy, uninteresting, require
longer attention span and don’t employ smartphone sensors
These platforms are generally driven by financial incentives alone
Fail to utilize the power of online newcomers using smartphones
mCrowd
Crowdsourcing framework keeping smartphone users in mind
Focused on creating entertainment driven environments
Implementation
Creating gamified incentive based mobile crowdsourcing marketplace
Conducting MCQ based quizzes in flipped classroom setting
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Introduction
What is crowdsourcing
Approach to solve problems by posing them to an open crowd
Reward is offered for solutions meeting some quality threshold
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Crowdsourcing in industry
Company initiative Goal
Amazon Mechanical turk general-purpose
Burda Media BurdaStyle fashion design (sewing)
CNN iReport news reporting
Eli Lilly InnoCentive scientific R & D
Facebook Translations language translation
Google Image labeler image labeling
Netflix Netflix prize data mining
Clickworker Virtual workforce general-purpose
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Crowdsourcing across various disciplines
Company Area
Maven Consulting
99designs Design
KickStarter Funding
Trada SEO/SEM
TopCoder Engineering
iStockPhoto Photos
Voice123 Voice-overs
Zero Flaws Image manipulation
Gengo Translation
Casting Words Transcription
Innocentive Research & Development
Samasource Corporate social responsibility
ReTel Surveillance analytics
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Crowdsourcing and Ubiquitous computing
Framework Area
Ushahidi Non-profit software company that develops free and open-
source software for information collection, visualisation,
and interactive mapping: crowdsourcing for social activism
and public accountability
OpenDataKit Free and open-source set of tools which help organizations
author, field, and manage mobile data collection solutions
Eclipse Scout Context-aware development framework for mobile appli-
cations, runs on a mobile android device, and provides
applications with automatic sensing of context and envi-
ronment information
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Incentive based crowdsourcing
Giving incentives to users generates traction and attracts users
Has many issues (spam detection, fraudulent data, multiple profiles)
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Games with a purpose
Users tend to give honest responses when a gamified version of the
task is presented to them (Reference: https://goo.gl/L7BxBU)
User visits & retention are entertainment-driven (not incentive)
Method proven to be hugely successful in the past
Examples
ESP game
Phetch
Peekaboom
Verbosity
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Pervasive computing
More users are coming online, most of them using smartphones
These smartphones have inbuilt sensors capable of capturing rich
data in real-time
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Problem statement
Issues with existing solutions
No existing platform to cater tasks suitable for mobile phones
Most platforms require longer attention user spans, and are boring
Generally driven by financial incentives alone
Our solution - mCrowd
Framework to support shorter tasks, suited for mobile phones
Works for different type of response collection and input formats
Implementation
mCrowd to create incentive based mobile crowdsourcing marketplace
mCrowd to conduct MCQ based quizzes in flipped classroom setting
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Problem statement
Issues with existing solutions
No existing platform to cater tasks suitable for mobile phones
Most platforms require longer attention user spans, and are boring
Generally driven by financial incentives alone
Our solution - mCrowd
Framework to support shorter tasks, suited for mobile phones
Works for different type of response collection and input formats
Implementation
mCrowd to create incentive based mobile crowdsourcing marketplace
mCrowd to conduct MCQ based quizzes in flipped classroom setting
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Problem statement
Issues with existing solutions
No existing platform to cater tasks suitable for mobile phones
Most platforms require longer attention user spans, and are boring
Generally driven by financial incentives alone
Our solution - mCrowd
Framework to support shorter tasks, suited for mobile phones
Works for different type of response collection and input formats
Implementation
mCrowd to create incentive based mobile crowdsourcing marketplace
mCrowd to conduct MCQ based quizzes in flipped classroom setting
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd - Concept map
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd to create incentive based marketplace
Architecture
Server side - Javascript, PHP, Python to handle business logic
Client side - Used phonegap framework to create mobile app
Database - MongoDB (NoSQL based)
Features
Types of users - Requesters & Workers
Types of tasks - General purpose & location specific
Response categories - Data rating, validation & generation
Data type involved - Text, images & files
E-Mail and SMS alerts
User notifications, profile, facebook sign-in, levels, badges & wallet
Ability to add questions/manage responses singularly as well in bulk
with/without custom time limits
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd to create incentive based marketplace
Architecture
Server side - Javascript, PHP, Python to handle business logic
Client side - Used phonegap framework to create mobile app
Database - MongoDB (NoSQL based)
Features
Types of users - Requesters & Workers
Types of tasks - General purpose & location specific
Response categories - Data rating, validation & generation
Data type involved - Text, images & files
E-Mail and SMS alerts
User notifications, profile, facebook sign-in, levels, badges & wallet
Ability to add questions/manage responses singularly as well in bulk
with/without custom time limits
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Example tasks (Inexhaustive list)
Question Task type
Do you like this image? Data validation
Is this image profane? Data validation
Rate this image out of 5 stars Data rating
Which of these 2 designs is better? Data validation
Which of these 2 designs is better and why? Data entry
Is the sentiment in this tweet, positive or negative? Data validation
Rate the sentiment in this tweet Data rating
Is this comment suitable for an e-commerce website? Data validation
Can you write a better definition for crowdsourcing Data entry
Help rate Campus Restaurant Data rating
Upload a photo of a red object around you Data generation
Are these two images that of Campus restaurant? Data validation
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
App-Server interaction
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Task-Data type interactions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Logical view
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Process view
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Deployment view
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Scenarios
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Worker signup/login
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Worker sample tasks
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Worker profile
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Requester
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Requester individual task upload
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd in flipped classroom setting
Working
Using mCrowd to conduct online MCQ based quizzes
These quizzes can act as precursors to lecture (revision medium,
homework) with optional weightage
Features
Quiz is available only for preset duration of time
Students can continue quiz where they left
The student needs to necessarily give the quiz in the classroom
(IP-checks)
Teacher gets graded response sheet at the end of the quiz
Future work
Can be extended to other types of questions as well
Gamification by “karma” points in forums for user participation
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd in flipped classroom setting
Working
Using mCrowd to conduct online MCQ based quizzes
These quizzes can act as precursors to lecture (revision medium,
homework) with optional weightage
Features
Quiz is available only for preset duration of time
Students can continue quiz where they left
The student needs to necessarily give the quiz in the classroom
(IP-checks)
Teacher gets graded response sheet at the end of the quiz
Future work
Can be extended to other types of questions as well
Gamification by “karma” points in forums for user participation
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
mCrowd in flipped classroom setting
Working
Using mCrowd to conduct online MCQ based quizzes
These quizzes can act as precursors to lecture (revision medium,
homework) with optional weightage
Features
Quiz is available only for preset duration of time
Students can continue quiz where they left
The student needs to necessarily give the quiz in the classroom
(IP-checks)
Teacher gets graded response sheet at the end of the quiz
Future work
Can be extended to other types of questions as well
Gamification by “karma” points in forums for user participation
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Student
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Snapshots - Teacher
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Outline
1 Thesis in a slide
2 Introduction
3 Background & Related work
4 Problem statement
5 mCrowd in incentive based marketplace setting
6 mCrowd in Classroom setting
7 Conclusions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Conclusions
Conceptualization of mCrowd
Created a framework (mCrowd) for incentive based mobile
crowdsourcing
Implemented two applications based on mCrowd
Working with phonegap - keep most of the data on client side itself,
efficient usage of LocalStorage variables to maintain sessions
Tips for avoiding spam
Be specific
Don’t be too cheap
Have a way of verifying the results
Weigh your options
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Future Work
Using social media connect, rewarding active players, using lives in a
game can lead to better user experience.
Enabling HTML5 canvas for mobile users can lead to a new form of
data collection - data collection by user drawings
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Questions
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions
Thank you!
Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing

More Related Content

Similar to arpitjain10327145thesispresentation9thjuly-150817073104-lva1-app6892

Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology managementsmumbahelp
 
Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology managementsmumbahelp
 
IRJET - Recommendations Engine with Multi-Objective Contextual Bandits (U...
IRJET -  	  Recommendations Engine with Multi-Objective Contextual Bandits (U...IRJET -  	  Recommendations Engine with Multi-Objective Contextual Bandits (U...
IRJET - Recommendations Engine with Multi-Objective Contextual Bandits (U...IRJET Journal
 
Achieving product market fit in startup context - The-state-of-practices and ...
Achieving product market fit in startup context - The-state-of-practices and ...Achieving product market fit in startup context - The-state-of-practices and ...
Achieving product market fit in startup context - The-state-of-practices and ...Anh Nguyen Duc
 
Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology managementsmumbahelp
 
Poster ECIS 2016
Poster ECIS 2016Poster ECIS 2016
Poster ECIS 2016Rui Silva
 
10 Questions on Crowdsourcing for Enterprise IT - Appirio
10 Questions on Crowdsourcing for Enterprise IT - Appirio10 Questions on Crowdsourcing for Enterprise IT - Appirio
10 Questions on Crowdsourcing for Enterprise IT - AppirioMark Longstaff
 
Relevant multimedia question answering
Relevant multimedia question answeringRelevant multimedia question answering
Relevant multimedia question answeringvembuking
 
NISI Introductie Continuous Delivery 3.0
NISI Introductie Continuous Delivery 3.0NISI Introductie Continuous Delivery 3.0
NISI Introductie Continuous Delivery 3.0Garm Lucassen
 
saftey chip in the market for smart watch
saftey chip in the market for smart watchsaftey chip in the market for smart watch
saftey chip in the market for smart watchTathagatMaurya3
 
Bath Spa Keynote T&L Dat
Bath Spa Keynote T&L DatBath Spa Keynote T&L Dat
Bath Spa Keynote T&L Datandyramsden
 
Samuel_Wayfinding_Metropolia_2015
Samuel_Wayfinding_Metropolia_2015Samuel_Wayfinding_Metropolia_2015
Samuel_Wayfinding_Metropolia_2015dbelgi
 
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...EI Design
 
Job-to-be-done theory to practice : Ch4 Process
Job-to-be-done theory to practice : Ch4 ProcessJob-to-be-done theory to practice : Ch4 Process
Job-to-be-done theory to practice : Ch4 ProcessPRADA Hsiung
 
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBus
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBusCorso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBus
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBusAlessandro Longo
 
On building more human query answering systems
On building more human query answering systemsOn building more human query answering systems
On building more human query answering systemsINRIA-OAK
 
Webinar Agile Presentation V.1.0
Webinar Agile Presentation V.1.0Webinar Agile Presentation V.1.0
Webinar Agile Presentation V.1.0fhios
 
I am sharing 'MP' with you.pptx for project
I am sharing 'MP' with you.pptx for projectI am sharing 'MP' with you.pptx for project
I am sharing 'MP' with you.pptx for projectVeerGohel1
 
MIL_Current and Future Trends Media and Information (1).pptx
MIL_Current and Future Trends Media and Information (1).pptxMIL_Current and Future Trends Media and Information (1).pptx
MIL_Current and Future Trends Media and Information (1).pptxCRISTYMAEDETALO
 
User Experience Versus Marketing
User Experience Versus MarketingUser Experience Versus Marketing
User Experience Versus MarketingIRJET Journal
 

Similar to arpitjain10327145thesispresentation9thjuly-150817073104-lva1-app6892 (20)

Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology management
 
Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology management
 
IRJET - Recommendations Engine with Multi-Objective Contextual Bandits (U...
IRJET -  	  Recommendations Engine with Multi-Objective Contextual Bandits (U...IRJET -  	  Recommendations Engine with Multi-Objective Contextual Bandits (U...
IRJET - Recommendations Engine with Multi-Objective Contextual Bandits (U...
 
Achieving product market fit in startup context - The-state-of-practices and ...
Achieving product market fit in startup context - The-state-of-practices and ...Achieving product market fit in startup context - The-state-of-practices and ...
Achieving product market fit in startup context - The-state-of-practices and ...
 
Mi0040 technology management
Mi0040  technology managementMi0040  technology management
Mi0040 technology management
 
Poster ECIS 2016
Poster ECIS 2016Poster ECIS 2016
Poster ECIS 2016
 
10 Questions on Crowdsourcing for Enterprise IT - Appirio
10 Questions on Crowdsourcing for Enterprise IT - Appirio10 Questions on Crowdsourcing for Enterprise IT - Appirio
10 Questions on Crowdsourcing for Enterprise IT - Appirio
 
Relevant multimedia question answering
Relevant multimedia question answeringRelevant multimedia question answering
Relevant multimedia question answering
 
NISI Introductie Continuous Delivery 3.0
NISI Introductie Continuous Delivery 3.0NISI Introductie Continuous Delivery 3.0
NISI Introductie Continuous Delivery 3.0
 
saftey chip in the market for smart watch
saftey chip in the market for smart watchsaftey chip in the market for smart watch
saftey chip in the market for smart watch
 
Bath Spa Keynote T&L Dat
Bath Spa Keynote T&L DatBath Spa Keynote T&L Dat
Bath Spa Keynote T&L Dat
 
Samuel_Wayfinding_Metropolia_2015
Samuel_Wayfinding_Metropolia_2015Samuel_Wayfinding_Metropolia_2015
Samuel_Wayfinding_Metropolia_2015
 
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...
5 Killer Examples : How to Use Microlearning Based Training Effectively - EI ...
 
Job-to-be-done theory to practice : Ch4 Process
Job-to-be-done theory to practice : Ch4 ProcessJob-to-be-done theory to practice : Ch4 Process
Job-to-be-done theory to practice : Ch4 Process
 
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBus
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBusCorso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBus
Corso Interazione Uomo Macchina e Sviluppo Applicazioni Mobile - GoBus
 
On building more human query answering systems
On building more human query answering systemsOn building more human query answering systems
On building more human query answering systems
 
Webinar Agile Presentation V.1.0
Webinar Agile Presentation V.1.0Webinar Agile Presentation V.1.0
Webinar Agile Presentation V.1.0
 
I am sharing 'MP' with you.pptx for project
I am sharing 'MP' with you.pptx for projectI am sharing 'MP' with you.pptx for project
I am sharing 'MP' with you.pptx for project
 
MIL_Current and Future Trends Media and Information (1).pptx
MIL_Current and Future Trends Media and Information (1).pptxMIL_Current and Future Trends Media and Information (1).pptx
MIL_Current and Future Trends Media and Information (1).pptx
 
User Experience Versus Marketing
User Experience Versus MarketingUser Experience Versus Marketing
User Experience Versus Marketing
 

arpitjain10327145thesispresentation9thjuly-150817073104-lva1-app6892

  • 1. mCrowd: A framework for incentive based mobile crowdsourcing Arpit Jain Thesis supervisor: Dr. T.V Prabhakar Department of Computer Science & Engineering IIT Kanpur July 13, 2015
  • 2. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 3. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 4. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Thesis in a slide Reasoning Current incentive based crowdsourcing platforms are not customized for smartphone users Most “tasks” on such platforms are lengthy, uninteresting, require longer attention span and don’t employ smartphone sensors These platforms are generally driven by financial incentives alone Fail to utilize the power of online newcomers using smartphones mCrowd Crowdsourcing framework keeping smartphone users in mind Focused on creating entertainment driven environments Implementation Creating gamified incentive based mobile crowdsourcing marketplace Conducting MCQ based quizzes in flipped classroom setting Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 5. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 6. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Introduction What is crowdsourcing Approach to solve problems by posing them to an open crowd Reward is offered for solutions meeting some quality threshold Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 7. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Crowdsourcing in industry Company initiative Goal Amazon Mechanical turk general-purpose Burda Media BurdaStyle fashion design (sewing) CNN iReport news reporting Eli Lilly InnoCentive scientific R & D Facebook Translations language translation Google Image labeler image labeling Netflix Netflix prize data mining Clickworker Virtual workforce general-purpose Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 8. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Crowdsourcing across various disciplines Company Area Maven Consulting 99designs Design KickStarter Funding Trada SEO/SEM TopCoder Engineering iStockPhoto Photos Voice123 Voice-overs Zero Flaws Image manipulation Gengo Translation Casting Words Transcription Innocentive Research & Development Samasource Corporate social responsibility ReTel Surveillance analytics Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 9. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Crowdsourcing and Ubiquitous computing Framework Area Ushahidi Non-profit software company that develops free and open- source software for information collection, visualisation, and interactive mapping: crowdsourcing for social activism and public accountability OpenDataKit Free and open-source set of tools which help organizations author, field, and manage mobile data collection solutions Eclipse Scout Context-aware development framework for mobile appli- cations, runs on a mobile android device, and provides applications with automatic sensing of context and envi- ronment information Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 10. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 11. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Incentive based crowdsourcing Giving incentives to users generates traction and attracts users Has many issues (spam detection, fraudulent data, multiple profiles) Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 12. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Games with a purpose Users tend to give honest responses when a gamified version of the task is presented to them (Reference: https://goo.gl/L7BxBU) User visits & retention are entertainment-driven (not incentive) Method proven to be hugely successful in the past Examples ESP game Phetch Peekaboom Verbosity Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 13. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Pervasive computing More users are coming online, most of them using smartphones These smartphones have inbuilt sensors capable of capturing rich data in real-time Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 14. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 15. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Problem statement Issues with existing solutions No existing platform to cater tasks suitable for mobile phones Most platforms require longer attention user spans, and are boring Generally driven by financial incentives alone Our solution - mCrowd Framework to support shorter tasks, suited for mobile phones Works for different type of response collection and input formats Implementation mCrowd to create incentive based mobile crowdsourcing marketplace mCrowd to conduct MCQ based quizzes in flipped classroom setting Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 16. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Problem statement Issues with existing solutions No existing platform to cater tasks suitable for mobile phones Most platforms require longer attention user spans, and are boring Generally driven by financial incentives alone Our solution - mCrowd Framework to support shorter tasks, suited for mobile phones Works for different type of response collection and input formats Implementation mCrowd to create incentive based mobile crowdsourcing marketplace mCrowd to conduct MCQ based quizzes in flipped classroom setting Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 17. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Problem statement Issues with existing solutions No existing platform to cater tasks suitable for mobile phones Most platforms require longer attention user spans, and are boring Generally driven by financial incentives alone Our solution - mCrowd Framework to support shorter tasks, suited for mobile phones Works for different type of response collection and input formats Implementation mCrowd to create incentive based mobile crowdsourcing marketplace mCrowd to conduct MCQ based quizzes in flipped classroom setting Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 18. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd - Concept map Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 19. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 20. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd to create incentive based marketplace Architecture Server side - Javascript, PHP, Python to handle business logic Client side - Used phonegap framework to create mobile app Database - MongoDB (NoSQL based) Features Types of users - Requesters & Workers Types of tasks - General purpose & location specific Response categories - Data rating, validation & generation Data type involved - Text, images & files E-Mail and SMS alerts User notifications, profile, facebook sign-in, levels, badges & wallet Ability to add questions/manage responses singularly as well in bulk with/without custom time limits Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 21. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd to create incentive based marketplace Architecture Server side - Javascript, PHP, Python to handle business logic Client side - Used phonegap framework to create mobile app Database - MongoDB (NoSQL based) Features Types of users - Requesters & Workers Types of tasks - General purpose & location specific Response categories - Data rating, validation & generation Data type involved - Text, images & files E-Mail and SMS alerts User notifications, profile, facebook sign-in, levels, badges & wallet Ability to add questions/manage responses singularly as well in bulk with/without custom time limits Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 22. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Example tasks (Inexhaustive list) Question Task type Do you like this image? Data validation Is this image profane? Data validation Rate this image out of 5 stars Data rating Which of these 2 designs is better? Data validation Which of these 2 designs is better and why? Data entry Is the sentiment in this tweet, positive or negative? Data validation Rate the sentiment in this tweet Data rating Is this comment suitable for an e-commerce website? Data validation Can you write a better definition for crowdsourcing Data entry Help rate Campus Restaurant Data rating Upload a photo of a red object around you Data generation Are these two images that of Campus restaurant? Data validation Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 23. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions App-Server interaction Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 24. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Task-Data type interactions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 25. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Logical view Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 26. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Process view Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 27. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Deployment view Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 28. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Scenarios Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 29. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Worker signup/login Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 30. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Worker sample tasks Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 31. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Worker profile Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 32. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Requester Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 33. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Requester individual task upload Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 34. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 35. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd in flipped classroom setting Working Using mCrowd to conduct online MCQ based quizzes These quizzes can act as precursors to lecture (revision medium, homework) with optional weightage Features Quiz is available only for preset duration of time Students can continue quiz where they left The student needs to necessarily give the quiz in the classroom (IP-checks) Teacher gets graded response sheet at the end of the quiz Future work Can be extended to other types of questions as well Gamification by “karma” points in forums for user participation Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 36. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd in flipped classroom setting Working Using mCrowd to conduct online MCQ based quizzes These quizzes can act as precursors to lecture (revision medium, homework) with optional weightage Features Quiz is available only for preset duration of time Students can continue quiz where they left The student needs to necessarily give the quiz in the classroom (IP-checks) Teacher gets graded response sheet at the end of the quiz Future work Can be extended to other types of questions as well Gamification by “karma” points in forums for user participation Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 37. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions mCrowd in flipped classroom setting Working Using mCrowd to conduct online MCQ based quizzes These quizzes can act as precursors to lecture (revision medium, homework) with optional weightage Features Quiz is available only for preset duration of time Students can continue quiz where they left The student needs to necessarily give the quiz in the classroom (IP-checks) Teacher gets graded response sheet at the end of the quiz Future work Can be extended to other types of questions as well Gamification by “karma” points in forums for user participation Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 38. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Student Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 39. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Snapshots - Teacher Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 40. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Outline 1 Thesis in a slide 2 Introduction 3 Background & Related work 4 Problem statement 5 mCrowd in incentive based marketplace setting 6 mCrowd in Classroom setting 7 Conclusions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 41. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Conclusions Conceptualization of mCrowd Created a framework (mCrowd) for incentive based mobile crowdsourcing Implemented two applications based on mCrowd Working with phonegap - keep most of the data on client side itself, efficient usage of LocalStorage variables to maintain sessions Tips for avoiding spam Be specific Don’t be too cheap Have a way of verifying the results Weigh your options Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 42. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Future Work Using social media connect, rewarding active players, using lives in a game can lead to better user experience. Enabling HTML5 canvas for mobile users can lead to a new form of data collection - data collection by user drawings Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 43. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Questions Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing
  • 44. Thesis in a slide Introduction Background Problem statement Marketplace setting Classroom setting Conclusions Thank you! Arpit Jain (10327145) mCrowd: A framework for incentive based mobile crowdsourcing