System design and development update
We're near the end of the 1st year of the grant and system development is progressing well, albeit a bit slower than we'd hoped. The high level system design is set and we've been implementing functionality in a series of sprints. By the end of the current sprint, we should have a basic system in place, allowing us to import email messages, import human annotations of some of the data from an Atlas-ti file, learn a model and apply the model to additional data. The final piece will be an interface to allow a human coder to correct the machine-applied annotations. Of course, once this is done, there'll still be a lot to do: clean up and enhance the interface, experiment with ML settings for best performance, import of other kinds of files, better reporting, etc. But it's exciting to see things start to come together.