November 7th, 2.15pm, the Thesis Defense: "A tool driven approach to Mechanical Turk user experiments" by Gobie Nanthakumar will take place at HCC Lab.

Abstract

Online experiments are becoming more and more popular in the field of Human-Computer-Interaction (HCI). One of the reasons for this is that a large sample can be collected with little effort, thus reducing costs. Amazons Service Mechanical Turk (MTurk) is one of the most popular crowdsourcing platforms and enables the distribution of tasks to workers. MTurk is used by the HCC research group, which has received promising results. Currently, however, there is still no workflow to perform experiments using MTurk, from creation to management of the experiments. MTurk does not currently offer the possibility to do this via a graphical interface, but only via the command line (AWS shell). The aim of this work was to develop a tool that facilitates the access and execution of MTurk experiments. The main feature was to bypass the AWS shell and replace it with the tool. The tool has been implemented as a web application, with Vue.js in the frontend, Node.js in the backend and MongoDB as the database. The requirements are based on semi-structured interviews with HCC researchers who have already used MTurk. The resulting prototype was finally tested and evaluated using the Teaching method and the existing MTurk experiment for Recoin.

The defence will be held in german.

First assesor: Prof. Dr. Claudia Müller-Birn
Second assesor: Prof. Dr. Marian Margraf

Location: HCC Lab, Königin-Luise-Straße 24/26, room 120