The term Crowdsourcing comes from Crowd and Outsourcing. Outsourcing is giving a task to do over the internet to a large group of people, instead of hiring someone to do the job physically. It helps companies outsource a job in the form of an open call to a large group of people.
With the internet, communities are able to form out of interest. What was once done in the context of a hierarchy can now be done in a community. The goal is to share knowledge. One main example of crowdsourcing is Wikipedia, but there are a wide variety of forms.
Crowdsourcing is interesting because with technology, the public’s capacity and skills are improving. As an example, photography was first exclusive for professionals who had very specific instruments. However, with the development of new cameras and technologies in the industry, more and more people were able to get into the domain. Also, editing softwares allowed for amateurs to gain skills and come closer to professionals. The internet has created a whole new community that is quite easy to join.
The idea of crowdsourcing is not new. Since forever, people have asked for help from others. Collaboration is at the root of human interaction. For example, in 1858, the first English dictionary was created thanks to the helps pf volunteers that helped putting it together.
Now, with the boom of the internet, everyone of is connect and reach people has become very easy. There are many different types of crowdsourcing such a crowdfunding (= raising capital) or crowd creativity which is related to medias.
This workshop focused on one particular type of Crowdsourcing: asking others to perform micro tasks in exchange for a fee. This is done on virtual crowdsourcing platforms such as Amturk and Figure Eight.
First of all, we saw examples of crowdsourcing tasks for digital humanities. For example, we examined models that were created to extract information from tweets related to disasters. Also, another example was to show pictures to the public and asking them to specify the gender and the race of the people in the images. Then, the information was shared to analyze discrimination on some platforms.
Secondly, the instructor demonstrated to us how to crowd source a micro-task on one of these platforms. The workflow of crowd sourcing is simple. We are the requesters. We have a task we want to achieve. The platform has people working. From these people, we have to do some recruitment by specifying some characteristics of the kind of workers we want. After this, we must specify and finish the task. Most of the time, these people are paid most of the time. However, most of them cannot be trusted. Thus, there are a lot of safety mechanisms such as result validation. Unless contributors, the money is not paid.
Then, it was our turn to practice what we learned. It started by brainstorming. We split into groups and thought about some tasks that we wanted to achieve. These tasks are micro tasks. They should be quick and simple. My group and I came with the following idea: we wanted to ask contributors to analyze tweets and categorize them in one of 5 categories related to problems in Lebanon: Air pollution, Garbage Crisis, Water, Electricity and Traffic.
First, we entered Figure eight, which is a platform for crowdsourcing. We then uploaded our data and specified the general title of the task. This title should be eye opening and tell the contributor everything about the task. Furthermore, we had to give the contributor some information about how to perform the job. The guidance is very important. Finally, we had to give some examples of the type of answers.
Overall, this experience was really eye opening on the multitude facets o Digital Humanities. The conference was very well organized and I enjoyed talking with the people that were present. This particular workshop about Crowdsourcing gave me an overview of a really interesting practice and I will definitely re use the knowledge I gained.
Humanities Digitalised 