Re: [eclipsecon-na-program-committee] Talks selection process

[Martin Lippert <mlippert@xxxxxxxxx>]

Hey!

I totally like the way Achim does this for ECE and I am happy to adopt this for NA as well.
But I think I would like to put more emphasis on the community voting… :-)

Just an idea: why not (more or less blindly) accept the top 10 talks of the community voting by default?
And then continue with our voting for the rest of the submissions?

And maybe do the same for the early-bird selection?
Select 2 spots via the community votes (top 2) and we select the remaining three spots (in case we pick 5)

I like the idea because it is not only asking the community for input, it involves the community and makes it part of the PC… :-)

What do you think?

Cheers,
-Martin







> Am 26.10.2015 um 12:46 schrieb Mikael Barbero <mikael@xxxxxxxxxxx>:
> 
> Hi again, 
> 
> The way I would like to do both selections is close to what Achim did for ECE. Here is a rough draft. 
> 
> Early birds
> 
> By Nov 17, 3pm CET, I would like you to send me your 5 favorite talks by email. I will compile them in a spreadsheet along with community voting and send it to you. We will meet later on a call to talk about the submissions which have a similar number of votes and decide of the early birds selection. 
> 
> 
> For the final selection
> 
> Before the first call, I will divide the submissions into groups and assign talks/tutorials to pairs of you. You will have to vote +1/0/-1:
> 
> +1 means you really like it, 
> 0 means it is not off topic but is not a talk you're fond of, 
> -1 means off topic, obvious sales pitch
> 
> As each talk will receive a pair of (+1/0/-1), it can get the following results with meanings:
> 
> +1/+1: 	we don't want to miss it
> +1/0: 	probably a good talk
> 0/0: 		somewhat interesting, 
> -1/-1: 	will be rejected on the first call if no one disagree
> -1/0: 	the -1 will need to be commented, if we all agree, it will be rejected
> -1/+1: 	we need to talk about these ones
> 
> We will have a first call to talk about the bold values above.
> 
> Then we will have to vote on every not already rejected talks and tutorials. Vote will be from from 1 to 5. I will compile all of these votes with means and std deviations along with community voting. We will meet again for a second call. We will discuss the proposals on which we don't agree, on which the community doesn't agree and on which we don't agree with the community! 
> 
> You will have the opportunity after this call to change your vote. I will compile the changes, and we will meet again for a third call where we set the the average value above which talks will accepted by default. We will also talk about how much the community voting will be accounted for. My proposal is to integrate it for 20% of the result. We are 12 on the PC (excluding Julien as he won't vote for standard program talks, he will lead the IoT day committee), so the community voting will count for ~2,5 people. This is only a proposal and it can (will) change depending on the number of votes we get from the community. 
> 
> Finally, we will meet for a fourth and last call to decide about the talks slightly above or below our expectation level. We will also check that all topics are well represented. 
> 
> Cheers,
> Mikael
> 
> 
> 
> 
> _______________________________________________
> eclipsecon-na-program-committee mailing list
> eclipsecon-na-program-committee@xxxxxxxxxxx
> To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> https://dev.eclipse.org/mailman/listinfo/eclipsecon-na-program-committee