The purpose of the software is to extract information from a large number of documents, and then merge this information together into a knowledge base. We think this could be pretty great because it would support:
- Conflict detection and resolution. In the process of combining information from many sources into a single knowledge base, the system should be able to mark conflicts of information. That would indicate an inconsistency or controversy between the documents, which could be flagged for further investigation.
- Naturally queryable aggregate knowledge. We anticipate being able to build a query interface that is a natural extension of this system: just run the query through the extraction process and compare the result for consistency with the knowledge base. This would make the system into a “dissonance engine,” useful for opposition research or the popping of filter bubbles.
I should say that neither of us knows exactly what we are doing. But Dave’s almost got his PhD in human syntax so I think we’ve got a shot at building a sweet parser. What’s more, we’ve got the will and plan. It will be open source, of course, and we’re eager for collaborators.
We have one problem: We don’t know what to call it. I can’t even make the GitHub account for our code until we have a good name. And until then we’ll be sending Python scripts to each other as email attachments and that will never get anywhere. Please help us. Tell us what to name our project. If we use your name, we’ll do something awesome for you some day.
Scratch that. We’re calling it Bluestocking. The GitHub repo is here.