Pybossa

field_project_name	Pybossa
field_project_purpose		Support citizen science activities:	Computing: Somewhat Thinking: Somewhat Sensing: Somewhat Gaming: Somewhat
field_home_page_URL
Total number of projects (estimate)
Total individual participants (estimate)
field_project_start_date
field_comments
pyBossa system (like Crowdcrafting.org) has three main components: • The BOSSA core • A task presenter, i.e. a HTML5 frontend that will load the task data into the DOM of a web page. Task Presenters can be either HTML5/JavaScript code that sits in the pyBossa server or they can be made with any language communicating through the API. • A task creator, i.e. a script that will upload tasks for the application to the BOSSA core. There are two approaches for creating pyBossa applications: • Using the web interface of a pyBossa server • Using the REST API You can create applications that use existing templates with different types of data. Else, you must learn how to create applications with Python. You will have to install a virtual Python environment in your machine, code, create the application that then can be "pushed" into a pyBossa server.	Completion level	Low

Documented projects in this wiki:

Sorry, the PDF print version needs yet to be done ....

Thinking: Computing: Sensing: Gaming:

pyBossa templates

pyBossa templates are reusable applications templates that implement popular crowd crafting problems. Templates include: a task creator, a task presenter, a tutorial, an application description, etc.

As of April 2013, there are five templates available on the crowdsourcing.org server. You can adapt these to your own needs in various ways.

Template	Task template	Description
EpiCollect Plus Project	EpiCollect Plus	Interface to an EpiCollect project
Geo-coding	Urban Park Task	Allows users to tag a park in a list of cities of the whole world. The result is then shown in a 3D Globe (rendered with WebGL)
Image Pattern Recognition	Flickr Person Finder	Uses the Flickr web service as the source of the data, pulls a picture from a Flick Feed and then asks the user a question with several possible answers.
Transcribing documents	PDF transcription	An application that allows transcribing PDF documents (e.g. old texts)
Sound Pattern Recognition	SoundCloud	...

In addition, there are two special templates allowing to import data from CSV files or Google Drive Spreadsheets. In order to use that data, you will have to write the code for a presenter.

Template	Task template	Description
CSV	Public link to a CSV file	?
Google Drive Spreadsheet	Import a Google Drive Spreadsheet	?

All PyBossa templates are available in GitHub repository, e.g. you can add these to your own pyBossa server or modify the code and push to a pyBossa server.

Data exports

Both tasks and task runs can be exported either in CSV or JSON format for analysis.

Adapting the built-in Image pattern recognition template

This example will show how to adapt the built-in Image pattern recognition using the public pyBossa sever.

Example that you only can view once, unless you connect from a different IP. It is really just a test, e.g. only includes three pictures of the 3D printers I own ....

• http://crowdcrafting.org/app/3D_printers_demo/

Below, we summarize the steps.

Create an account and an application

• Create an account on http://crowdcrafting.org/
• Click on the Create button on top menu to create an application

Fill in 8 fields that will describe the application. You are allowed to change these later.

You can find your applications in your "personal" pull down menu.

Select an existing application template

In the Settings for your application (personal pull-down menu -> My applications):

• Click on Import tasks and select the Image Pattern Recognition template

You now will see a link to a Google spreadsheet and that you can substitute by your own. Therefore, do not use this link, create your own as explained below.

• Go to Google Drive and log in. If you don't have a Coogle account you will have to create one.
• Open the this spreadsheet

• Create a copy and share it on your own google drive. With the open file, use menu File->Share.
• Replace all the pictures and maybe also the questions. Do not change the headings, since the presenter will use these.

col1 = question

col2 = thumbnail picture

col3 = credits page

col4 = full picture

... of course you can link to any picture from any site. E.g. we took ours from this wiki:

Create/Modify the task presenter

Pick the Image recognition task (the others will not work with the template)

You now can change the number and the phrasing of the answers. As you can see below, we added an answer:

<!-- If the user clicks this button, the saved answer will be value="maybe"-->
<button class="btn btn-answer" 
        value='Maybe'><i class="icon icon-white icon-comment-alt"></i>Maybe</button>

Adapting the buttons PyBossa uses the Twitter bootstrap CSS framework.

• icon-comment-alt refers to a special font that was made for this framework. In principle, you can use any of these font names from this list.
• Same principle for the button color styles, e.g. btn_answer would be light grey. See the Twitter bootstrap buttons list

Original fragment:

<div id="answer"> <!-- Start DIV for the submission buttons -->
<!-- If the user clicks this button, the saved answer will be value="yes"-->
<button class="btn btn-success btn-answer" value='Yes'><i class="icon icon-white icon-thumbs-up"></i> Yes</button>
<!-- If the user clicks this button, the saved answer will be value="no"-->
<button class="btn btn-danger btn-answer" value='No'><i class="icon icon-white icon-thumbs-down"></i> No</button>
<!-- If the user clicks this button, the saved answer will be value="NotKnown"-->
<button class="btn btn-answer" value='NotKnown'><i class="icon icon-white icon-question-sign"></i> I don't know</button>
</div><!-- End of DIV for the submission buttons -->

New fragment:

<div id="answer"> <!-- Start DIV for the submission buttons -->
<!-- If the user clicks this button, the saved answer will be value="yes"-->
<button class="btn btn-success btn-answer" value='Yes'><i class="icon icon-white icon-thumbs-up"></i> Yes</button>
<!-- If the user clicks this button, the saved answer will be value="no"-->
<button class="btn btn-danger btn-answer" value='No'><i class="icon icon-white icon-thumbs-down"></i> No</button>
<!-- If the user clicks this button, the saved answer will be value="maybe"-->
<button class="btn btn-answer" value='Maybe'><i class="icon icon-white icon-comment-alt"></i>Maybe</button>
<!-- If the user clicks this button, the saved answer will be value="NotKnown"-->
<button class="btn btn-answer" value='NotKnown'><i class="icon icon-white icon-question-sign"></i> I don't know</button>
</div><!-- End of DIV for the submission buttons -->

The result looks like this:

Adapted PyBossa application

A simple example reusing existing code

Draft

1. Create an account and/or an application as above

Create an application

• Install a Python virtual environment. This way you can be sure that dependencies will remain ok (not overwritten by system upgrades, etc.)

• Start from an existing template, e.g. from the Application tutorial or from an application you could download from the GitHub repository

.....

Create and push the application to the pyBossa server

python createTasks.py -u http://crowdcrafting.org -k API-KEY

• Select your application (personal pull-down menu on top -> My Applications
• Open your application.

Links

Official documentation

Documentation is available on docs.pybossa.com/. For example:

• PyBossa Overview
• Build with PyBossa
• Quick overview: Creating your own Application
• Step by step tutorial on creating an Application (more advanced use)

• Crowdcrafting.org. Official PyBossa server that hosts many projects. Anyone can create his/her own.

Example applications

• Digital Humanitarian Response: Moving from Crowdsourcing to Microtasking

Other

• Open Knowledge Labs has posts concerning pyBossa, e.g. about new features

BIBLIOGRAPHY