|
|
 |
 |
 |
 |
 |
 |
 |
Vogoo - Recommendation Engine and Collaborative Filtering
With Vogoo you can :
- recommend items that match your customers'tastes
- include a "people who liked this item also liked..." feature on your website using item-based collaborative filtering
- predict users's ratings with the Slope One algorithm
- compute tastes'similarities between users and create communities based on them
and all it requires is just a few lines in your pages scripts.
 |
Vogoo PHP Pro v2.2 |
Vogoo PHP Pro 2.2 is now available and adds the ability to limit the number of items returned by item-based CF functions.
With Vogoo PHP Pro v2.2 you can get up-to-date recommendations whenever a rating is added/updated, thanks to realtime correlation calculations. Vogoo PHP Pro v2.2 also includes optimizations to the recommendation and prediction engines for medium to large datasets.
Click here for more information on Vogoo PHP Pro
|
Vogoo PHP Lib v2.2 |
This release adds the ability to limit the number of items returned by item-based CF functions.
Vogoo PHP includes 2 item-based and 1 user-based recommendation engines.
It's been designed with ease-of-use in mind. With Vogoo PHP you can add professionnal collaborative filtering functions to your website in minutes.
Vogoo PHP Lib v2.2 is a free PHP library licensed under the terms of the GNU GPL v2 or later.
Vogoo PHP Pro v2.2 - New !
Vogoo PHP Pro is a proprietary version of the library, highly optimized for item-based Collaborative Filtering pre-computations.
Performance tests have shown an improvement of up to 20x in execution speed for pre-computation scripts when compared to the GPL version.
Version 2.2 adds the ability to limit the number of items returned by item-based CF functions.
Vogoo PHP Pro requires a MySQL database.
Purchase Vogoo PHP Pro from the Pro page.
Vogoo PHP v2.x new features
- Indexes and constraints added in the DB
- DB schema and code change for item-based CF
- New statistics module (thanks Jan)
- Code reorganization that improves modules consistency
- Most returned results now use associative arrays
- SQL code fix that works with DB modules provided in the db/ area (tested on MySQL and PostreSQL only)
- Fixed bugs in the visitor module
Other features
- item-based collaborative filtering using the "people who like this item also like..." concept
- Slope one engine based on Daniel Lemire's work. This engine can predict users'ratings.
- user-based collaborative filtering
- computation of accurate similarities between users
- item-to-item recommendation functions (if you like these items, you'll also like...) that do not require any knowlege of the user's tastes
- tailored recommendations via filtering
- multiple item categories
- members and visitors/guests ratings management
- automatic ratings based on purchases and page views
- let users specify when they are not interest in an item
-
- real-time targeted ads or messages
Main releases
03/29/2008 - v2.2: limit the number of items returned by item-based CF functions
02/28/2008 - v2.0.1 and Pro v2.1.1: Fixed a bug in Slope calculation (thanks Alex)
02/14/2008 - PHP Pro v2.1.1: optimized item-based engines, up-to-date recommendations when ratings are updated/changed
09/23/2007 - v2.0 : statistics module, new DB schema, associative arrays in returned results, bug fixes, improved code consistency
03/10/2006 - v1.6 : item-to-item recommendation functions
10/23/2005 - v1.5 : item-based Collaborative Filtering engines
08/22/2005 - v1.0 : targeted ads
04/21/2005 - Initial release of Vogoo PHP LIB v0.8 on Sourceforge and Freshmeat: user-based Collaborative Filtering and visitors functions.
References
Vogoo products are used on the following websites :
If you use Vogoo on your website and would like us to add a link in this section, please send us an e-mail.
Vogoo Express and Vogoo PHP Pro are Coyright © 2007 Stéphane Droux.
Vogoo PHP Lib is released under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 2 of the License, or at your option,
any later version.
|
|
|
| |
|