Tags

, , , , , , ,

In this post, we will focus on two important concepts: motivation and crowdsourcing. The latest is widely used in projects related to Digital Humanities because it enables us to perform a task using the large connected crowd available through the Internet. But, why does it work? What motivates crowds to spend time online to contribute to a project’s success, be it transcription of old texts or massive data collection? We will try to answer these questions comparing three examples of projects that have used crowdsourcing to achieve their goals.

What is crowdsourcing?

First of all it is important to thoroughly understand the concept of crowdsourcing. The first person to give a definition of the crowdsourcing was Jeff Howe, in 2006 on the “Crowdsourcing blog”. According to him, “crowdsourcing represents the act of a company or institution taking a function once performed by employees and outsourcing it to an undefined (and generally large) network of people in the form of an open call”. Furthermore, Enrique Estellés-Arolas and Fernando Gonzalez-Ladron-de-Guevara from the department of Management of the technical University of Valencia in Spain offered to give a complete definition of the crowdsourcing in 2012 after having studied more than 40 different definitions of the world “crowdsourcing” in the literature. They published their resulting definition in the article “Towards an integrated crowdsourcing definition” [5]:

Crowdsourcing is a type of participative online activity in which an individual, an institution, a non-profit organization, or company proposes to a group of individuals of varying knowledge, heterogeneity, and number, via a flexible open call, the voluntary undertaking of a task. […] The user will receive the satisfaction of a given type of need, be it economic, social recognition, self-esteem, or the development of individual skills, while the crowdsourcer will obtain and utilize to their advantage that what the user has brought to the venture, whose form will depend on the type of activity undertaken.

Figure 1: Word cloud for the “Crowdsourcing” term. For more information about word cloud, see this blogpost. Source.

We now understand that crowdsourcing is an online process that enables small communities of scientists or researchers to undertake big projects thanks to the participation of a lot of different people, with their own useful skills. It is interesting to note that the authors mention in the definition itself that crowdsourcing is a voluntary undertaking, and that they mention the satisfaction that the crowd performing the task will likely experience. There are a lot of different tasks that can be performed by the crowd. The help of the crowd is usually required for endeavors that are based on the analysis of a huge amount of historical data, or to create new database. Historically, one of the first examples of tasks performed by a crowd is the creation of the Oxford English Dictionary in the 18th century. More recently, we can obviously mention the example of the large, online open-encyclopedia known as Wikipedia. To get more information about crowdsourcing, just have a look at this video.

Three recent examples of tasks using crowdsourcing in the field of the Digital Humanities 

The first example is based on the paper “Crowdsourcing Performing Arts History with NYPL’s ENSEMBLE” written by Doug Reside [1]. The New York Public Library (NYPL) has a big collection of theater, dance or music programs from performances given in the US from the end of the Civil War, and estimates their number at roughly 1 million. In 2013, they began to digitalize these documents in order to transcribe them by using a crowdsourcing approach. It is easy to understand that, with such a large amount of documents, asking for the help of the crowd is a really good idea. They launched their project, named ENSEMBLE on a website. On this site, there is a list of digitalized paper-programs, and the task of the crowd (which you can be part of; so, go take a look!) is to recognize then transcribe the title, the date, the name of the theater, etc., i.e. all the relevant information available on the program. The final aim is to structure the database of the programs thanks to the transcribed texts.

The next example is a typically Swiss one, because the task of this crowd-based project is to develop a smartphone application to collect information about all the different Swiss German dialects. It is based on the paper “Swiss voice App: A smartphone application for crowdsourcing Swiss German dialect app” by Marie-José Kolly et al. [2]. The goal of these researchers from the Universities of Zurich and Geneva is to develop an app called “Voice Äpp”, which will be the continuation of the “Dialäkt App”. The latter is already capable of recognizing the region of origin of a Swiss German speaker when the user selects (among a list that appears on the app) the way (s)he pronounces 16 given words. The “Voice Äpp” will achieve the same goal but, instead of asking the user to choose the “correct” pronunciation in a list of words, it will ask him/her to directly pronounce these words using the smartphone’s microphone (ASR technology). This task is not easy to implement and the ASR system needs training. Thankfully, on the “Dialäkt App”, it is also possible to record his/her own voice, which then constitutes a new crowdsourcing database that can be used to train the ASR system for the forthcoming “Voice Äpp”.

Capture d’écran 2014-10-16 à 11.01.2

Figure 2: Mockup of the playful part of the Voice Äpp, from the Univeristy of Zürich. Source.

The last example [3] relates to an educational experiment. Two members of the Rhetoric faculty of the Univeristy of Iowa (Matt Gilchrist and Tom Keegan), in association with two librarians (Jen Wolfe and Kelly McElroy), tried to introduce the concept of crowdsourcing in a general undergraduate rhetoric course. At the initiative of some libraries, one website was launched in 2011: the DIY History. This website is also based on crowdsourcing and is in fact a “public digital humanities resource” [3] which can also be used for educational purposes. In fact, it presents you (the crowd) the opportunity to help them transcribe digitalized manuscripts from the Civil War. Furthermore, the two teachers incorporated this DIY website directly into their rhetoric course. The goals for the students were firstly to transcribe an old manuscript document (based on the DIY website) and then to write a blog post that contained some additional information about the rhetoric, and the historical context of this document.

Through these three examples we now understand the power of the crowd and its utility in the context of Digital Humanities. It can help scientists for tasks that seem to be huge for a small scientific community, but become easier to perform with the collaboration of hundreds of people through the Internet. The crowd can also be used to construct big database by inviting the connected people, for example, to record their voices through the “Dialäkt App” or the “Voice Äpp”.

A short introduction about the motivation

Motivation

Figure 3: The two orientations of the motivation. Source.

When a person decides to take part in a project, (s)he always has a certain motivation inside her/him. It appears clear that motivation is the process that allows somebody to be moved to do a given task. In their paper “Intrinsic and Extrinsic Motivations: Classic Definitions and New Directions” [4], Richard M. Ryan and Edward L. Deci from the University of Rochester argue that there is not necessarily different degrees of motivation, but rather two types of motivation, that they called “orientation of the motivation”. Thus, these two orientations of motivation are called “intrinsic” or “extrinsic”. These authors define intrinsic motivation as

the doing of an activity for its inherent satisfactions rather than for some separable consequence.

In this case, it’s our curiosity, and our appreciation of fun and challenging undertakings that draw us to the project. Humans are naturally “designed” to have all these qualities, especially to allow them to develop their social capabilities. It can be seen as natural motivation, where rewards or pressure aren’t necessary to push somebody to accomplish a task. On the other hand, people are not always intrinsically motivated to do something, even though we all have natural motivation to do tasks, moved by our curiosity. So, the extrinsic motivation relates to

an activity […] done in order to attain some separable outcome

(same authors, [4]). We can understand extrinsically motivated task as the instrument used by the person in order to receive some external reward (for example, work for a salary). A good TED video allows us to fully grasp the difference between these two orientations of motivation. It’s a Daniel Pink’s (journalist) speech.

Now the question is: are the people taking part of a crowdsourcing project generally moved by an intrinsic or an extrinsic motivation?

The motivation for crowdsourcing in Digital Humanities

Two of our crowdsourcing examples (the ENSEMBLE and DIY) show that crowdsourcing in the context of Digital Humanities is often used to transcribe old documents in a digital form, i.e. transcribe an old text in order to record it in numerical form. Generally, these documents are historical documents, which depict the life of our ancestors, and help us to understand how they lived.

 The authors of the paper “Crowdsourcing in the Curiculum” [3] report that DIY web-platform “has been an unqualified success” and that many volunteer users of the website report a “connection with the original writers of the documents” when transcribing a document. Furthermore, they say: “These people [the ancient authors of the manuscript] come alive; you come to share their hopes, their fears, their everyday concerns. You do become absorbed in the people, and losing one of them is like losing a relative or a friend” (from [3]). We can already see that the crowd is motivated by its curiosity for the past. Moreover, students of the rhetorical course report also diverse origins of their motivation to do this transcription: “producing meaningful work”; “encountering new information”; “learning new skills”; “fun” and “working collaboratively to achieve a big goal” (from [3]). Other students’ feedbacks are, for instance: “nobody else had read the letter my project was based on”, “it was a challenge”, “I’m pretty proud”, “homework actually can be fun”. With all these examples, it is obvious that the connected crowd and the students have intrinsic motivation. Nobody will give them money or recognition for their tasks, and yet they accomplished it! It’s quite clear that they achieved their transcription moved by the fun, the challenge and the connection established with ancient authors.

In the “Voice Äpp” example, what motivates the user to record her/his voice in the app? In this case, if the user records her/his voice, (s)he can “learn about characteristics of her/his own voice in a playful way”, like the speech rate, the pitch, the articulation and have visual perception (from [2]). For the authors, the app should obviously be very interactive to be sure that the app is really interesting for the user and, moreover, be sure that (s)he will record in turn her/his voice. So, the user records his/her voice in order to access to this interactive part of the app and begin to play with the available options (Figure 2). The user is once more moved by intrinsic motivation: if I record my voice, I can satisfy my curiosity because I can play with the different options.

Our last example (the ENSEMBLE project, [1]) shows us the importance of the web-interface for the motivation of crowds. Indeed, the authors speak about a very low participation for the project. They argue that it can be explained by the difficulty of the task (which is not only an easy transcription) or some problems with the interface itself. More interesting, they note a small detail. In fact, when someone of the crowd has transcripted a document, they require a validation of the transcription before incorporating it in the database; this proofreading is done by other users. Initially, they chose to show the level of agreement and the user could see it directly after the submission of the post. The problem was that some people “saw that the system had a low “degree of confidence” in the work they had just submitted (since no one else had yet “agreed” with them) misunderstood what they were being told and felt either insulted or disheartened” (from [1]). This detail is clearly in contradiction with intrinsic motivation because the initial fun and challenging task is immediately blown away by this system of validation. The intrinsic motivation that the user could find by performing this task is now switched off, and that could explain the low participation in this project. The authors removed this functionality. We therefore understand the importance of the interface because if it’s not as fun as possible, or as interactive as desired by the user (in order to be sure that we will touch her/his intrinsic motivation), nobody will take part in the project, especially since there are no extrinsic factors of motivation.

Conclusions

We have seen different concepts that are important in order to be sure that the crowd will take part in a project based on crowdsourcing. In fact, in this context, people are moved by intrinsic motivation factors. As we’ve just explored, the simple fact of becoming a part of a big project, which includes the treatment of old texts or the constitution of a big database in a playful way, is generally sufficient to tickle the natural curiosity of the humans, because we are intrinsically social subjects, moved by challenges and fun. So, the projects related to Digital Humanities that are opened to crowdsourcing pertain to digitalized information that people will find meaningful, and that will motivate them to take part in the project. So, at what level can the designers of crowdsourcing web-platforms or apps act to promote the motivating factors contained in the task itself?

It is then important that the designers of a web-platform or an app think thoroughly about its attractiveness, and its simplicity of use. Indeed, if the user is lost when (s)he’s visiting the website, it will rapidly kill her/his intrinsic motivation. It’s also important to avoid some functionalities that can be misunderstood by the user to be sure that (s)he won’t receive a negative feedback after (s)he has posted her/his contribution. Obviously, the designers can’t give motivation, but they could and should construct a well-designed platform in order to insure that the use of the website won’t annihilate the intrinsic motivation of the crowd for a task in the field of the Digital Humanities.

Thus, it’s really important to realize that the user will generally never receive a reward if (s)he is part of a crowdsourcing-based Digital Humanities project. So we have to reach the crowd, tickle its curiosity, and be sure that the task will be as fun as possible to promote natural intrinsic motivation. The only way to do that is to design attractive, easy-to-use and fun web-platforms.


 REFERENCES

3 papers from Digital Humanities session ’14 (EPFL): 

[1] Doug Reside, “Crowdsourcing Performing Arts History with NYPL’s ENSEMBLE“, Long Paper presented the 11th of July 2014 at Digital Humanities session ’14.

[2] Kolly, Leemann, Dellwo, Goldman, Hove & Almajai, “Swiss Voice App: A smartphone application for crowdsourcing Siss German dialect data“, Short Paper presented the 9th of July 2014 at Digital Humanities session ’14.

[3] Gilchrist, Wolfe, McElroy, Keegan, “Crowdsourcing in the Curriculum: Engaging Undergraduates through Collaborative Manuscript Transcription“, presented the 10th of July 2014 at Digital Humanities session ’14.

Others sources: 

[4] Richard M. Ryan and Edward L. Deci, “Intrinsic and Extrinsic Motivations: Classic Definitions and New Directions“, University of Rochester, in Contemporary Educational Psychology 25, 54–67, 2000.

[5] Enrique Estellés-Arolas and Fernando Gonzalez-Ladron-de-Guevara, Towards an integrated crowdsourcing definition”, Department of Management, Technical University of Valencia, Spain in the Journal of Information Science, XX(X) pp.1-14, 2012.

Advertisements