Facing the challenges that were not remotely foreseeable at the start of the journey: The Thinkwise team joined the UNLOCK program with the purpose of creating an online tool that allowed people to separate the wheat from the chaff, on the current most “channelizing channel”, the internet. This blog gives you some insights into their journey.
The Wikimedia Accelerator UNLOCK has made it possible for us to work as a team. So, we had to learn to leave room for the individuals to develop as a glued coherent team, while respecting each other’s visions, tempo, personal needs, aptitudes and motivations. The idea has been adapted so that people incorporated to the program could develop their own interests and strengths. We would like to highlight that running the first sprint alongside with the other 4 UNLOCK project teams was a good catalyst for this – creating a healthy team cookie cutter environment.
How do we manage to come down from tons of ideas to be more concrete? How to put something-into-a-Chrome-extension, and generate algorithms, train datasets, refactor existing open-source code etc. Which ideas do we want to start with or leave aside? We have a wide range of ideas that varies from “argument mining” (detecting arguments in the text the user is reading), crowdsourcing universal annotations, pre-highlighting units of information and many more. All the discussions around these ideas have consumed the most resources and time. Let us highlight here the invaluable support of our coach Eileen Wagner who helped us sail through this no-winds sea. It’s all about decisions but in order to make them, you have to build a minimum foundation. But how to?
User testing seems to be fundamental to address the above questions, especially finding the different personas, and therefore a “market” (we don’t like this word actually). We greatly benefitted from the coaching in order to better understand “how to product” and conduct proper user research and testing. We have interviewed a preliminary batch of people that have served us to identify problems and interests that people might have. We would like to clarify what is useful or not to the target users and, at the same time, learn how people are compiling information. We are interested in the question: What criteria do you use when you highlight texts?
In the past two months, we have prototyped a minimalistic Chrome extension that allows you to assign up to 4 different categories to whatever text you highlight. These 4 categories are the universal attributes that define knowledge quality and will match our framework. We are discussing what we should integrate next in the prototype based on the results from the user research. Some of the elements to be integrated are: the framework (highlighting criterias), the possibility of retrieving claims (part of arguments) in webpages or the possibility of using the “clauses” to select highlights.
We are currently running a second round of user research that will hopefully help us to answer the following question: How such highlighter could be used by researchers (one of our main target users)? How are we going to find a “market fit”? What sustainability/business model are we going to have to not only survive but become an alternative tool in the long term? We care about becoming an alternative tool that really opens some doors to the way things are done.
