Machine learning combined with human annotations to grow a neural network
Following Andrzej Marczewski’s gamification model R.A.M.P. we demonstrate the purpose of user contribution. First, our onboarding explains that every contibution helps other researchers to find interconnected topics of their research. Secondly, every annotation is rewarded with a micro-animation of growing connections.
Autonomy is an important factor for users to deliberately interact with the content. Therefore we let them choose a page on a topic of their interest at the end of the onboarding. Furthermore, they can choose on their own which words they want to annotate.
By a means of tunneling users are asked follow-up questions on annotations they made. These questions follow a What - Where - When - Why logic, which form extended connections to other topics. To users this seems like a quiz while at the same time we grow the body of knowledge.
In the beginning, we broadened our perspective on the project by identifying our clients’ objectives, engaging with experts and users, and analyzing the competitors.
In the second phase, we defined the main emphasis within the problem space and identified the goal and the users of our project by means of the gathered information.
Afterwards, we indivudally generated several sketches influenced by our findings from research.
Together, we evaluated each of the ideas , iterated on them, and chose one concept would then be transferred to the digital space.
In this phase, we envisioned our concept through a digital prototype to compare different solutions and validate various aspects in the following phase.
In the validation phase, we gathered feedback and valuable insights from our users. These findings served together with the following ideation phase as a foundation to redefine the problem space.