[
https://issues.apache.org/jira/browse/FLINK-4613?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15516571#comment-15516571
]
ASF GitHub Bot commented on FLINK-4613:
---------------------------------------
Github user gaborhermann commented on a diff in the pull request:
https://github.com/apache/flink/pull/2542#discussion_r80253321
--- Diff: docs/dev/libs/ml/als.md ---
@@ -49,6 +49,18 @@ By applying this step alternately to the matrices $U$
and $V$, we can iterativel
The matrix $R$ is given in its sparse representation as a tuple of $(i, j,
r)$ where $i$ denotes the row index, $j$ the column index and $r$ is the matrix
value at position $(i,j)$.
+An alternative model can be used for _implicit feedback_ datasets.
+These datasets only contain implicit feedback from the user
+in contrast to datasets with explicit feedback like movie ratings.
+For example users watch videos on a website and the website monitors which
user
+viewed which video, so the users only provide their preference implicitly.
+In these cases the feedback should not be treated as a
+rating, but rather an evidence that the user prefers that item.
+Thus, for implicit feedback datasets there is a slightly different
+minimalization problem to solve (see [Hu et
al.](http://dx.doi.org/10.1109/ICDM.2008.22) for details).
+Flink supports both explicit and implicit ALS,
+and the choice between the two can be set in the parameters.
+
--- End diff --
Okay, I added
"The implementation is based on the Apache Spark implementation of implicit
ALS."
and referred to the relevant file in the Spark codebase.
> Extend ALS to handle implicit feedback datasets
> -----------------------------------------------
>
> Key: FLINK-4613
> URL: https://issues.apache.org/jira/browse/FLINK-4613
> Project: Flink
> Issue Type: New Feature
> Components: Machine Learning Library
> Reporter: Gábor Hermann
> Assignee: Gábor Hermann
>
> The Alternating Least Squares implementation should be extended to handle
> _implicit feedback_ datasets. These datasets do not contain explicit ratings
> by users, they are rather built by collecting user behavior (e.g. user
> listened to artist X for Y minutes), and they require a slightly different
> optimization objective. See details by [Hu et
> al|http://dx.doi.org/10.1109/ICDM.2008.22].
> We do not need to modify much in the original ALS algorithm. See [Spark ALS
> implementation|https://github.com/apache/spark/blob/master/mllib/src/main/scala/org/apache/spark/ml/recommendation/ALS.scala],
> which could be a basis for this extension. Only the updating factor part is
> modified, and most of the changes are in the local parts of the algorithm
> (i.e. UDFs). In fact, the only modification that is not local, is
> precomputing a matrix product Y^T * Y and broadcasting it to all the nodes,
> which we can do with broadcast DataSets.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
