[Noisebridge-discuss] call for collaboration: nb-discuss data mining

[Mike Schachter]

Just a reminder that tonight at 7:30pm in the Church
room we're having a brainstorming session on how to
data-mine the shit out of noisebridge-discuss:

1) Distribution and Overview of the Preprocessed Data
2) Analyzing the Social Network of nb-discuss (PageRank Overview)
3) Topic-based Analysis (TF*IDF, Latent Dirichlet Allocation Overview)
4) Brainstorming Session



On Fri, May 18, 2012 at 11:34 AM, Mike Schachter
<mschachter at eigenminds.com> wrote:
> Noisebridge Machine Learning
> https://www.noisebridge.net/index.php?title=Machine_Learning
> ----------------------------------
>
> Hey there everyone,
>
> A while ago Jake suggested we code up an algorithm to
> do drama detection on the nb-discuss list. Jared D then
> went on to start grabbing data from nb-discuss and we're
> going to begin the laborious process of preprocessing it
> into a form that would be good to do some wholesome
> old fashioned machine learning.
>
> We're going to hold an ML meeting in the Church room
> on Thursday, 5/24 at 7:30pm to discuss what direction
> we can take the data, and need your help! In the meeting
> we'll distribute the preprocessed data and collaboratively
> brainstorm ways to have fun with it. We'll give some brief
> overviews of text-processing ideas like TF*IDF, LDA (Latent
> Dirichlet Allocation), and PageRank/network analysis
> algorithms.
>
> Anybody with interest or experience in text data mining
> is encouraged to come out and share your expertise and
> ideas with the group.
>
> See you Thursday!
>
>  mike