The Knowledge Factory is an interactive machine learning and data analytics environment (also known as human-in-the-loop machine learning or AI) that provides tight integration between machine learning and knowledge acquisition from experts. This pioneering research program included one of the few detailed experimental demonstrations that human-in-the-loop machine learning can outperform both autonomous machine learning and an expert specifying rules without the assistance of machine learning.
Software
The Knowledge Factory is a Macintosh System 4 to 7 application.
The Knowledge Factory is best run under Mac OS 7.6. It runs successfully under the Basilisk II Mac emulator on Windows.
TKF.sea is a self extracting archive. Double click on the TKF.sea icon and follow the instructions.
The archive contains the following files:
- The Knowledge Factory - application
- manual.hdr, manual.dat, manual,evl - data files used within the Manual.pdf
- hypo.hdr, hypo.dat, hypo.evl - data files containing 100 hypothyroid cases
- ReadMe - a text file containing features of the application not documented withing the Manual.pdf
- Manual.pdf - Adobe's Portable document format requires the Adobe Acrobat reader
The Knowledge Factory:
- requires no special installation. Simply double click on the The Knowledge Factory icon and the system becomes operational.
- will run on any Macintosh computer operating under System 4.1 or higher so long as sufficient RAM is available.
- is initially configured to request 1000K when it is run under Multifinder. This should normally be sufficient for projects examining up to 4000 example cases.
If The Knowledge Factory displays a message during operation stating that it is unable to complete an action due to a shortage of memory, you should increase the application memory size by:
- saving the current state of the project;
- quitting from The Knowledge Factory;
- selecting the The Knowledge Factory icon;
- choosing Get Info from the Finder’s File Menu; and
- entering the desired value in the Preferred Size box.
Publications
[bibtex file=interactiveML.bib template=publist-norel format=my-format sort=year order=desc process_titles=0]