Jump to content

Research talk:Are the bots really fighting/Work log/2017-03-01

Add topic
From Meta, a Wikimedia project coordination wiki

Wednesday, March 1, 2017

[edit]

Tsvetkova et al. point out that there's a lower bot revert rate in German Wikipedia but a very high rate in Portuguese Wikipedia. I think most of the "reverts" they find are just bots keeping the redirect graph clean. So maybe German Wikipedia has few redirects and Portuguese Wikipedia has many. Let's find out!

mysql:research@analytics-store.eqiad.wmnet [ptwiki]> select count(*) from page where page_namespace = 0 and page_is_redirect;
+----------+
| count(*) |
+----------+
|   736419 |
+----------+
1 row in set (0.52 sec)

mysql:research@analytics-store.eqiad.wmnet [ptwiki]> select count(*) from page where page_namespace = 0;
+----------+
| count(*) |
+----------+
|  1695025 |
+----------+
1 row in set (1.04 sec)

mysql:research@analytics-store.eqiad.wmnet [ptwiki]> SELECT 736419/1695025;
+----------------+
| 736419/1695025 |
+----------------+
|         0.4345 |
+----------------+
1 row in set (0.00 sec)

OK so ptwiki has 43.4% of it's main namespace pages redirecting somewhere else.

What about German?

mysql:research@analytics-store.eqiad.wmnet [dewiki]> select count(*) from page where page_namespace = 0 and page_is_redirect;
+----------+
| count(*) |
+----------+
|  1379216 |
+----------+
1 row in set (0.89 sec)

mysql:research@analytics-store.eqiad.wmnet [dewiki]> select count(*) from page where page_namespace = 0;
+----------+
| count(*) |
+----------+
|  3416846 |
+----------+
1 row in set (1.92 sec)

mysql:research@analytics-store.eqiad.wmnet [dewiki]> select 1379216/3416846;
+-----------------+
| 1379216/3416846 |
+-----------------+
|          0.4037 |
+-----------------+
1 row in set (0.00 sec)

Oh.. 40.37% redirects. I wonder how the number of page moves per page looks.


mysql:research@analytics-store.eqiad.wmnet [ptwiki]> select count(*) from logging where log_type="move" and log_namespace = 0;
+----------+
| count(*) |
+----------+
|   253247 |
+----------+
1 row in set (35.62 sec)

mysql:research@analytics-store.eqiad.wmnet [ptwiki]> SELECT 253247/1695025;
+----------------+
| 253247/1695025 |
+----------------+
|         0.1494 |
+----------------+
1 row in set (0.00 sec)

mysql:research@analytics-store.eqiad.wmnet [dewiki]> select count(*) from logging where log_type="move" and log_namespace = 0;
+----------+
| count(*) |
+----------+
|   503824 |
+----------+
1 row in set (2 min 30.25 sec)

mysql:research@analytics-store.eqiad.wmnet [dewiki]> select 503824/3416846;
+----------------+
| 503824/3416846 |
+----------------+
|         0.1475 |
+----------------+
1 row in set (0.00 sec)

15% vs 14.8%! Not a big difference. --EpochFail (talk) 23:45, 1 March 2017 (UTC)Reply