Hi All,

I’m very excited to say that the University of Maryland has executed an CCLA 
with Software Grant for the Apache Flagon project!

The Software Grant encompasses a refactor of the Apache Flagon Distill product, 
which has been deprecated for a few years now [1]  

This Grant includes ~150 new commits and thousands of new insertions (and a lot 
of deletions)

At UMD my team and I re-thought Distill as a Python Package (for distro through 
PyPI) [2] that allows users to:

- efficiently segment UserALE.js (or User Behavior Logs) data
- curate segments
- transform segments with logical operations (intersection, union, etc.)
- filter log data extracted from segments
- apply analytics (e.g., statistical, graphs) to data extracted from segments
- support graph-based visualization (funnel, sankey)

Additionally: 
- code is well documented [3]
- excellent working examples [4] 

Distill also provides examples for dashboards to visualize segments using both 
Apache Superset and Plotly/Dash

The original Distill product was tethered to a front-end (Tap), relied on 
rudimentary (and error-prone) processing of segments within the client. This 
made Distill difficult to maintain and required users to adopt tap, limiting 
their analytical use-cases with UserALE.js and Distill.

Overall, I think that Distill provides a far more scaleable product to engage 
(and expand) our development community. I think this will add real value to 
Apache Flagon.

Overhead/Actions prior to release:

- Documentation builds will need be adjusted for a new repo/branch
- Additional documentation for a few analytical functions (i.e., click-rate)
- Minor tweaks to simply code in a few functions (i.e., click-rate)
- Some restructuring of Repo to consolidate and organize examples
- a few additional README’s should be added (i.e., examples /dir)

IP Clearance

The Software Grant is signed by Felicia Metz, esq. who is an Associate Director 
at the University of Maryland’s, UMD Ventures office, which has authority to 
release UMD IP. That makes accepting this grant significantly easier from an IP 
Clearance perspective.

New Committers

The UMD executed CCLA includes 5 new committers and potentially new PPMC 
members.

Prior to initiating a VOTE on private@ I wanted to give the community an 
opportunity to chime in and ask questions. The New Committers listed are 
already subscribed to the dev list.

Let’s take 72 hours to DISCUSS prior to a VOTE.

Best,

Josh


[1] https://github.com/apache/incubator-flagon-distill 
<https://github.com/apache/incubator-flagon-distill>
[2] 
https://github.com/UMD-ARLIS/incubator-flagon-distill/tree/distill_toolkit_refactor
 
<https://github.com/UMD-ARLIS/incubator-flagon-distill/tree/distill_toolkit_refactor>
 
[3] 
https://incubator-flagon-distill.readthedocs.io/en/distill_toolkit_refactor/ 
<https://incubator-flagon-distill.readthedocs.io/en/distill_toolkit_refactor/> 
[4] 
https://github.com/UMD-ARLIS/incubator-flagon-distill/blob/distill_toolkit_refactor/examples/Segments_Demo.ipynb
 
<https://github.com/UMD-ARLIS/incubator-flagon-distill/blob/distill_toolkit_refactor/examples/Segments_Demo.ipynb>
 

Reply via email to