We wanted to give our all our Vici entries the opportunity to write about their submitted tools. Wikinizer held back on their post so they could write about their new release (which will be called MindGraph). So we’re happy to be able to share a post from Andrew Benedek as an early Christmas present! Enjoy!
WikiNizer is a next generation concept organization tool which supports personal knowledge work. It was built as a mobile first hybrid app for Android, using a simple in-house graph database implemented in SQLite. Two years ago we started to rebuild it (based on the Capapbility Graph below) as a web app with an OrientDB graph database.
We selected OrientDB because of its multi model support, preparing the way for the development of Conceptipedia – which supports collaborative knowledge work. In the end we found that the OrientDB graph model did not quite give us what we required, and so we built our own, requiring only cloud storage.
We targeted a number of cloud storage providers, from CloudME, OpenDrive to SME Storage. Eventually we have found Google Drive as a compelling proposition, and developed a way of using Google Docs which incorporated WikiNizer’s own simple semantic markdown into Personal Knowledge Graphs and back. With that in place we were ready, using Freebase, to integrate with Google’s Knowledge Graph, and decided to submit this version, called WikiNizeR Research, as an entry for the LinkedUp Challenge. (Please click here to see our Poster.)
We found that for persistence, collaboration, and creating a Personal Knowledge Graph, Google Docs was exceedingly useful. It enables us to use WikiNizer, even in its current early state, as our tool of choice for web research and conceptualization. A week before the deadline however, just after we committed ourselves to enter the competition, we received the feedback that without a direct manipulable graph editor we would have a hard time of getting users to understand how to use our Google Docs format. We had numerous Google Docs we could work on, and load into WikiNizer, but we did not have time to revise them, and were instead swamped with new development tasks, just for us to reach only the 80% of success by turning up at all. Still it was a valuable exercise, and we are grateful for the opportunity, and the feedback we had received. The main lesson we have learned is the importance of setting “low pass filters” on what is to be tackled in order to complete what is needed to get off the ground.
Over the past months we have added the most glaring missing features, such as integrating WikiNizeR and Freebase searches, and defining viable end to end workflows, that were previously incomplete. We are currently working on releasing the kernel of WikiNizer, which we now call MindGraph. It will shortly be launched at alpha.wikinizer.com.