The Duine Recommender Framework consists of java libraries with logic to calculate how interesting information items are for a user.
It uses a mixture of clever algorithms (for example collaborative filtering) as well as user profiles to predict the level of interest. And..... it can learn: users can provide feedback (ratings) that will be stored in the profiles and used by the algorithms to get even better predictions.... read more
At the end of January 2009 a new version of the Duine Recommender will be released. Duine is a collection of software libraries that allows developers to create prediction engines for their own applications. The software has a new architecture, greatly improving the pluggability of user profiles, prediction algorithms and caching. The new architecture makes maximum use of existing technologies like Hibernate and Spring Framework.... read more
We have just released version 0.1 of a demonstration application for the Duine Toolkit. The demo is meant for developers to use as a working example application that uses the Duine Toolkit.
The first open source version of The Duine Toolkit is now released. This is version 2.0.3 of the toolkit. The release consists of five packages: one containing all files, one containing only the sources, one containing the compiled jar version, one containing the documentation and one containing support libraries.
Please read the readme.txt or release notes for more details on this release.
We hope to release a test application and step-by-step walkthrough of how to setup that test application in May.
The first beta version (including sources) will be uploaded shortly.