data and text mining workshop the role of crowdsourcing anna noel-storr wellcome trust, london,...

23
Data and text mining workshop The role of crowdsourcing Anna Noel-Storr Wellcome Trust, London, Friday 6 th March 2015

Upload: kevin-jackson

Post on 29-Dec-2015

215 views

Category:

Documents


0 download

TRANSCRIPT

Data and text mining workshopThe role of crowdsourcing

Anna Noel-StorrWellcome Trust, London, Friday 6th March 2015

What is crowdsourcing?

“…the practice of obtaining needed services, ideas, or content by soliciting contributions from a large group of people, and especially from an online community, rather than from traditional employees…”

Image credit: DesignCareer

What is crowdsourcing?

Knowledge discovery

and management

Brabham’s problem focused crowdsourcing typology: 4 types

What is crowdsourcing?

Knowledge discovery

and management

Broadcastsearch

Brabham’s problem focused crowdsourcing typology: 4 types

What is crowdsourcing?

Knowledge discovery

and management

Broadcastsearch

Peer-vetted creative

production

Brabham’s problem focused crowdsourcing typology: 4 types

What is crowdsourcing?

Knowledge discovery

and management

Broadcastsearch

Peer-vetted creative

production

Distributed human

intelligence tasking

Brabham’s problem focused crowdsourcing typology: 4 types

What is crowdsourcing?

Knowledge discovery

and management

Broadcastsearch

Peer-vetted creative

production

Distributed human

intelligence tasking

Brabham’s problem focused crowdsourcing typology: 4 types

Micro-tasking: process

Breaking down large corpus of data into smaller units and distributing those units to a large online crowd

“the distribution of small parts of a problem”

Human computation

Humans remain better than machines at certain tasks: e.g. Identifying pizza toppings from a picture of a pizzae.g. “preventing obesity without eating like a rabbit”.ti. – autotag: Animal study

Tools and platformsWhat platforms and tools exist and how do they work?

Image credit: ThinkStock

The Zooniverse

“each project uses the efforts and ability of volunteers to help scientists and researchers deal with the flood of data that confronts them”

Classification and annotation

Galaxy Zoo

Operation War Diary

Health related evidence productionCan we use crowdsourcing to identify the

evidence in a more timely way?

- Known pressure point within the review production- Between 2000 and 5000 citations per new review, but can be much more- A not much loved task

Trial identification

The Embase project

Cochrane’s Central Register

of Controlled Trials:

CENTRAL

EmbaseCrowd

Embase auto

Step 2: Use a crowd to screen thousands of search results from Embase and feed the identified reports of RCTs into CENTRAL

How will the crowd do this?

Step 1: run a very sensitive search in the largest biomedical database for studies

The screening tool

Three choices

You are not alone!

(and you can’t go back)

Progress bar

Yellow highlights to indicate a likely RCT

Red highlights

The Embase project: recruitment

- 900+ people have signed-up to screen citations in 12 months- 110,000+ citations have been collectively screened

- 4,000 RCTs/q-RCTs identified by the crowd

Feb-14 Mar-14 Apr-14 May-14 Jun-14 Jul-14 Aug-14 Sep-14 Oct-14 Nov-14 Dec-14 Jan-15 Feb-15 Mar-150

100

200

300

400

500

600

700

800

900

1000

Number of Participants

Participants

Why do people do it?

Made it very easy to participate (and equally easy to stop!)

Gain experience(bulk up the CV)

Provide feedback: both to the individual and to

the community

Wanting to do something to contribute (healthcare is a strong hook)

(people are more likely to come back)

RCT RCT RCT

Reject Reject Reject

Unsure

CENTRAL

Bin

Resolver

How accurate is the crowd?

RCTReject Resolver

5%

Crowd accuracy

TP 1565

FP 9

FN 2

TN 2888

TP415

FP 5

FN 1

TN 2649

The Crowd:INDEXTEST

The Crowd:INDEXTEST

The Info specialist: REFERENCE STANDARD

The Info specialists: REFERENCE STANDARD

Validation 1Validation 2

Sensitivity: 99.9% Specificity: 99.7% Sensitivity: 99.8% Specificity: 99.8%

Enriched sample; blinded to crowd decision; dual independent screeners as reference standard

Enriched sample; blinded to crowd decision; single independent expert screener (me!) as reference standard; possibility of incorporation bias

Individual screener accuracy is also carefully monitored

How fast is the crowd?

Number of weeks

Jan 2014 Jul 2014 Jan 2015

6 weeks

5 weeks

2 weeks

More screeners and more screeners screening more quickly

Length of time to screen one month’s worth of records

More of the same, and more tasks

As the crowd becomes more efficient, we plan to do two things:1. Increase the databases we search – feed in more citations2. Offer other ‘micro-tasks’

Feed in more citations – from other databases Bin

Y

N

ScreenAnnotate, appraise

And in these tasks the machine plays a vital and complementary role…

e.g. is the healthcare condition Alzheimer’s disease? Y, N, Unsure

Perfect partnership

Machine driven probability + Collective human decision-making

It’s not one or the other, the ideal is both

In summary

• Effective method in large scale study identification

• Identify more studies, more quickly

• No compromise on quality or accuracy

• Offers meaningful ways to contribute

• Feasible to recruit a crowd• Highly functional tool• Complements data and text

mining

And enables the move towards the living review

Crowdsourcing: