Re: Item Based Collaborative Filtering
From: Peter Kim [MS] (peterkim_at_online.microsoft.com)
Date: 06/07/04
- Previous message: Jen: "Item Based Collaborative Filtering"
- In reply to: Jen: "Item Based Collaborative Filtering"
- Next in thread: Jen: "Thank you Re: Item Based Collaborative Filtering"
- Reply: Jen: "Thank you Re: Item Based Collaborative Filtering"
- Messages sorted by: [ date ] [ thread ]
Date: Mon, 7 Jun 2004 11:27:23 -0700
In SQL Server 2000 data mining, we support a nested table column to model
exactly what you just mentioned. Collaborative filtering on tens of
thousands of items are just no problem at all. The FAQ in
(http://groups.msn.com/AnalysisServicesDataMining/faq.msnw) has entries to
show how to use a nested table columns (in model definition and prediction
input) and how to perform a collaborative filtering task using
Microsoft_Decision_Trees.
-- Peter Kim This posting is provided "AS IS" with no warranties, and confers no rights. "Jen" <anonymous@discussions.microsoft.com> wrote in message news:EA1AB5A5-D902-4F40-B7D0-AE16C7E896E1@microsoft.com... > > Has anyone used SQL Server to perform Item Based Collaborative Filtering? > The concern is the columns per table limitations in computing the initial > reference table. The table schema for the computational algorithm is > item1...item1000+ columns and customers in the rows. Once the initial > computation is in place, SQL Server is fine. I'm thinking about using > horizontal partions items1-1000 table1, items1001-2000 table 2 etc...any > other approaches, insight, shared experiences are appreciated! > > Thanks. > Jen
- Previous message: Jen: "Item Based Collaborative Filtering"
- In reply to: Jen: "Item Based Collaborative Filtering"
- Next in thread: Jen: "Thank you Re: Item Based Collaborative Filtering"
- Reply: Jen: "Thank you Re: Item Based Collaborative Filtering"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
