Joey Frazee commented on SPARK-8696:

Is there still interest in this? The two use cases I've seen for this are (1) 
low latency, or near real-time topic generation -- imagine a dashboard or 
process depending on _always_ up-to-date topic dist. -- and (2) desire to do 
updates rather than fitting the entire dataset again because it's very large or 
very expensive to pre-process – though maybe merely having topic-word priors 
such as suggested in SPARK-9134 could be a good enough alternative for this 
second use case. I've seen both of those requirements appear in tandem.


> Streaming API for Online LDA
> ----------------------------
>                 Key: SPARK-8696
>                 URL: https://issues.apache.org/jira/browse/SPARK-8696
>             Project: Spark
>          Issue Type: New Feature
>          Components: MLlib
>            Reporter: yuhao yang
>            Priority: Major
> Streaming LDA can be a natural extension from online LDA. 
> Yet for now we need to settle down the implementation for LDA prediction, to 
> support the predictOn method in the streaming version.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org

Reply via email to