User:Quaid/Community metrics: Difference between revisions

Latest revision as of 20:00, 15 August 2012

I am interested in quantifying and analyzing data generated by the Fedora Project as a way to measure the health of the Fedora community. This idea is directly connected to existing efforts to do the same, such as statistics++.

As a way to learn-as-we-go and get myself a useful set of data about the health of the Fedora community, I'm going to focus on a few broad strokes of information. These broad strokes potentially comprise a number of granular study points - places to look, and metrics to pull on a regular basis.

The processes here - the what and how - are going to change over time, in response to whatever happens.

Trends

Usage

Direct downloads
Number of repo connections

Usage data and notes 2012-08-14

https://docs.google.com/spreadsheet/ccc?key=0Amg7gpK9wXxFdFhVM3ZZWXRiQXNOQXVpSnZycHdVTlE#gid=2

Search relevance

Evolve a set of parameters for search engine metrics
- Can we drive down to see what is being done with Fedora? I.e., more than just talked about, how is it used?

Search relevance data and notes 2012-08-14

Keywords are in this specific order so the first term sets the Search Volume Index. The first time is the one we are comparing against, presumably it is going to be the common or generic base. For example, compare "Linux, Fedora" sets Linux as 1.0, and the term Fedora scores based against that (either higher or lower.)

Keywords as comparisons for search trends:

Social media search for trends:

'Fedora' on SocialMention

Method

Look at each link for current score.
On the trend page, look for Export this page as a CSV file > CSV with fixed scaling
- Use this to get specific numbers for previous points in time

Discussions

Forums
Mailing lists

Discussions data and notes 2012-08-15

Need to setup new systems to track here:
- Start mlstats for Fedora
- Start Mailman subscriber counts in aggregate
- Start ThinkUp App for Fedora

Content

Blogs
Videos
Tutorials
Articles
Maker/doer
- Can we drive down to see what is being done with Fedora? I.e., more than just talked about, how is it used?

Content data and notes 2012-07-25

Sources

The Statistics page and its sources provide the initial #Usage trends.

@@ Line 13: / Line 13: @@
 ==== Usage data and notes 2012-08-14 ====
+* https://docs.google.com/spreadsheet/ccc?key=0Amg7gpK9wXxFdFhVM3ZZWXRiQXNOQXVpSnZycHdVTlE#gid=2
+=== Search relevance ===
+* Evolve a set of parameters for search engine metrics
+** Can we drive down to see what is being done with Fedora? I.e., more than just talked about, how is it used?
+==== Search relevance data and notes 2012-08-14 ====
 Keywords are in this specific order so the [http://www.google.com/intl/en/trends/about.html#13 first term sets the Search Volume Index]. The first time is the one we are comparing against, presumably it is going to be the common or generic base. For example, compare "Linux, Fedora" sets Linux as 1.0, and the term Fedora scores based against that (either higher or lower.)
@@ Line 37: / Line 46: @@
 # On the trend page, look for ''Export this page as a CSV file'' > ''CSV with fixed scaling''
 #* Use this to get specific numbers for previous points in time
-=== Search relevance ===
-* Evolve a set of parameters for search engine metrics
-** Can we drive down to see what is being done with Fedora? I.e., more than just talked about, how is it used?
-==== Search relevance data and notes 2012-07-25 ====
 === Discussions ===
@@ Line 50: / Line 52: @@
 * Mailing lists
-==== Discussions data and notes 2012-07-25 ====
+==== Discussions data and notes 2012-08-15 ====
+* Need to setup new systems to track here:
+** Start mlstats for Fedora
+** Start Mailman subscriber counts in aggregate
+** Start ThinkUp App for Fedora
 === Content ===