Review rating prediction/recommendation interface

[ghost]

(originally reported in Trac by @elehack on 2011-04-05 14:26:36)

The current prediction/recommendation interface for rating-based systems takes the user as a long ID and a SparseVector of their ratings. This may couple the interface too closely to LensKit, making bridges (#14, #15) to other systems more difficult to write. Further, it does not allow the prediction or recommendation method to use the timestamps of the user's ratings, or to deal with duplicate ratings.

Therefore, we should consider changing the interfaces to use Collection<Rating> as the rating input.

Note: This issue has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:43.807165+00:00, last updated: 2013-02-01T22:35:41.274590+00:00

[ghost]

ekstrand on 2011-06-15 21:32:12 said:

This is now done.

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:46.363675+00:00, last updated: None

[ghost]

Michael Ekstrand ekstrand@cs.umn.edu on 2011-06-15 20:36:08 said:

In [3675339fde95fb93a13f870e8ac44328cb329c06]:

Clean up naming and documentation of recommenders (refs #84, #47)

* Rename predictor-based recommenders
* Collapse Simple and Abstract predictor-based recommenders into single concrete classes
* Drop BasketRecommender

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:45.973053+00:00, last updated: None

[ghost]

ekstrand on 2011-06-13 20:47:48 said:

Still clearing up some naming on this.

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:45.588726+00:00, last updated: None

[ghost]

Michael Ekstrand ekstrand@cs.umn.edu on 2011-06-13 20:27:48 said:

In [8ed7a58f7af2da18ef2f02bc67e2876853b3ee5d]:

Revert rename to restore `DynamicRatingItemRecommender` and friends (refs #47).

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:45.234054+00:00, last updated: None

[ghost]

Michael Ekstrand ekstrand@cs.umn.edu on 2011-06-13 20:27:48 said:

In [d86e4631218ef82893dd2b2ba3a524f3c5c1db68]:

Merge recommender refactorying & impl (refs #47)

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:44.879356+00:00, last updated: None

[ghost]

kolb0128@umn.edu on 2011-06-13 20:27:48 said:

In [e4476be7a9fc89d9caa680ad57187cf1f64b6b4a]:

Review rating prediction/recommendation interface (closes #47).

* Removed the getPredictableItems method from all rating predictors.
* Added a getPredictableItems method to all rating recommenders.
* Revised AbstractRatingRecommender to include generic implementations of the getPredictableItems and the
        recommend(long, SparseVector, int, LongSet, LongSet) methods.
* Added getPredictableItems and  revised recommend(long, SparseVector, int, LongSet, LongSet) methods in
        ItemItemRatingRecommender and UserUserRatingRecommender.
* Added the static method isComplete to SparseVector to determine if any keys are mapped to Double.NaN
* Fixed UserUserRatingRecommender to properly deal with null candidate sets. (Closes #62)
* Wrote JUnit test cases for all recommend methods implemented by ItemIItemRatingRecommender and
        UserUserRatingRecommender.

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:44.495543+00:00, last updated: None

[ghost]

ekstrand on 2011-04-08 17:10:03 said:

If we do this, then we need to have a further BaselinePredictor interface that preserves the SparseVector methods.

Note: This comment has been automatically migrated from Bitbucket
Created by grouplens on 2013-02-01T21:55:44.125554+00:00, last updated: None