Recommendation Algorithms for tweets in C#

Question

I'm looking for an algorithm/ recommendation engine to "recommend" tweets based on rating of the content of the tweet:

From a data set of 1000 rated(1 to 5) tweets recommend tweets based on the rated tweets from another data set with say ... 10 000 000 tweets.

What we are looking for is something already implemented on C#/.net , we looked at Aforge but we are not sure which algorithm suits our purpose best.

Also we looked at Google Prediction Engine but it's paid and we are doing this just for fun.

So getting my ass downvoted: http://xkcd.com/1060/ – Joe May 26 '12 at 15:48 — Joe, May 26 '12 at 15:48

score 5 · Accepted Answer · edited May 23 '17 at 11:44

5

You are looking for collaborative filtering family of algorithms. Here you can find several options for C# implementations. Among them there's implementation for SlopeOne algorithm - quite simple, but still very efficient one.

edited May 23 '17 at 11:44

Community

1
1

answered May 27 '12 at 14:48

ffriend

27,562
13
91
132

score 4 · Answer 2 · answered Jan 16 '14 at 19:56

it seems Apache Mahout Recommendation framework ideally suits your task - Mahout is a mature, production-ready scalable solution for tasks like that. If you need good introduction into this engine "Mahout in Action" book could help.

Despite of the fact that Mahout recommendation engine is Java library, now it is possible to use its complete C# port named NReco Recommender

Recommendation Algorithms for tweets in C#

2 Answers2