crowdflower-package: crowdflower: R Client for the CrowdFlower API

In cloudyr/crowdflower: R Client for the CrowdFlower API

Description

R Crowdflower API Client

Details

The crowdflower package provides programmatic access to the Crowdflower API. Crowdflower is a prominent crowdsourcing platform that connects “customers” to a large universe of “channels”, where “contributors” complete small tasks in exchange for money. Unlike some other crowdsourcing platforms, Crowdflower can distribute work to multiple “channels”, enabling you to draw a diverse, international set of crowdworkers.

Crowdflower works through “jobs”, where one job can involve multiple “rows” or “units” of input data. For example, in a sentiment analysis task, a customer (you) might have 1000 tweets to be coded for sentiment. This would constitute one job with 1000 rows/units that will be completed by contributors.

The basic workflow is to create a job, add data to the job, configure the “channels” that the job will be distributed to, and to finally retrieve the results. All of these actions can be performed with the crowdflower package.

The first step is to create a Crowdflower requester account and retrieve an API key from your account. To begin working, load the crowdflower package and specify your API key as an environment variable. Within R, this can be set with: Sys.setenv("CROWDFLOWER_API_KEY" = "example12345apikeystring").

Users may also be interested in the MTurkR package, which provides access to the Amazon Mechanical Turk crowdsourcing platform.

Author(s)

Maintainer: Kenneth Benoit [e-mail]

Authors:

• Pablo Barberá [e-mail]

• Thomas J. Leeper [e-mail]

References

API Overview API Response Codes API Request Guide

See Also

job_create, job_channel_list, job_add_data, job_status, report_get

Examples

## Not run: 
# create new job
f1 <- system.file("templates/instructions1.html", package = "crowdflower")
f2 <- system.file("templates/cml1.xml", package = "crowdflower")
j <- job_create(title = "Job Title", 
                instructions = readChar(f1, nchars = 1e8L),
                cml = readChar(f2, nchars = 1e8L))

# add one or more channels
job_channel_add(j, "neobux")

# upload data
d <- data.frame(variable = 1:3)
job_add_data(id = j, data = d)

# launch the job
job_launch(j)

# monitor the job
job_status(j)

# get aggregated results once completed
report_get(j)

# get results
results_get(j)

# delete job
job_delete(j)

## End(Not run)

cloudyr/crowdflower documentation built on May 13, 2019, 8:20 p.m.