Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.php > #3180 > unrolled thread

Stats comp.lang.php (last 7 days)

Back to article view | Back to comp.lang.php

Contents

  Stats comp.lang.php (last 7 days) me <me@privacy.net> - 2011-09-19 07:21 +0000
    Re: Stats comp.lang.php (last 7 days) pittendrigh <sandy.pittendrigh@gmail.com> - 2011-09-24 06:01 -0700
      Re: Stats comp.lang.php (last 7 days) houghi <houghi@houghi.org.invalid> - 2011-09-24 16:41 +0200
        Re: Stats comp.lang.php (last 7 days) pittendrigh <sandy.pittendrigh@gmail.com> - 2011-09-24 09:54 -0700

#3180 — Stats comp.lang.php (last 7 days)

 "Caveat: Quantity is not necessarily a measure of Quality"
 
 
 Newsgroup.................: comp.lang.php
 Stats Were Taken..........: Mon, 19 Sep 2011 07:17:34 GMT
 Stats Begin...............: Mon, 12 Sep 2011 07:41:13 GMT
 Stats End.................: Sat, 17 Sep 2011 13:38:54 GMT
 Days......................: 7
 Total No. of Articles.....: 58
 Total No. of Characters...: 125798
 Total Volume..............: 122
 Messages Per Day..........: 8.3
 Characters Per Day........: 17971.1
 Average Daily Volume......: 17 kB
 Total Posters This Week...: 28
 Messages with Sigs........: 44.83%
 Original Content Rating...: 51.13%
 
 Top 10 Prolific Posters: Posts / Posts per Day / Percent Share
 ==============================================================
   1. Jerry Stuckle.....................................:  11  1.6 19.0%
   2. Peter H. Coffin...................................:   4  0.6  6.9%
   3. A.................................................:   3  0.4  5.2%
   4. .......123maza..../65/Chill155/...................:   3  0.4  5.2%
   5. =?Utf-8?B?Isobbhzhcm8gry4gvmljyxjpbyi=?=..........:   3  0.4  5.2%
   6. Denis McMahon.....................................:   3  0.4  5.2%
   7. Thomas 'Pointedears' Lahn.........................:   3  0.4  5.2%
   8. Bob...............................................:   3  0.4  5.2%
   9. Comp Lang Php.....................................:   3  0.4  5.2%
  10. Rataplanbumbum....................................:   3  0.4  5.2%
 
 Top 10 Bandwidth-Using Posters: kBytes / kBytes per Day / Percent Share
 =======================================================================
   1. Jerry Stuckle.....................................:  28  4.1 23.2%
   2. Peter H. Coffin...................................:   9  1.3  7.5%
   3. =?Utf-8?B?Isobbhzhcm8gry4gvmljyxjpbyi=?=..........:   8  1.3  7.3%
   4. Thomas 'Pointedears' Lahn.........................:   8  1.2  6.6%
   5. Bob...............................................:   5  0.8  4.7%
   6. Rataplanbumbum....................................:   5  0.8  4.3%
   7. A.................................................:   4  0.6  3.5%
   8. Denis McMahon.....................................:   4  0.6  3.5%
   9. Comp Lang Php.....................................:   4  0.6  3.4%
  10. Chuck Anderson....................................:   4  0.6  3.4%
 
 Top 10 Popular Threads: Posts / Posts per Day / Percent Share
 =============================================================
   1. best editor.......................................:  17  2.4 29.3%
   2. Social Bookmarks Script...........................:   7  1.0 12.1%
   3. PHP Developer.....................................:   6  0.9 10.3%
   4. PHP, MySQL and UTF-8?.............................:   5  0.7  8.6%
   5. Trying to decode text that is supposed to be ISO-8:   5  0.7  8.6%
   6. date parsing......................................:   3  0.4  5.2%
   7. see...............................................:   3  0.4  5.2%
   8. Notice: Undefined index: action error please help.:   2  0.3  3.4%
   9. Startup City 2011 (September 24th,Mumbai).........:   2  0.3  3.4%
  10. What would you like to see in most in a text edito:   2  0.3  3.4%
 
 Top 10 Bandwidth-Consuming Threads: kBytes / kBytes per Day / Percent Share
 ===========================================================================
   1. best editor.......................................:  29  4.2 23.9%
   2. Social Bookmarks Script...........................:  20  2.9 16.6%
   3. Trying to decode text that is supposed to be ISO-8:  12  1.8 10.3%
   4. PHP Developer.....................................:  10  1.5  8.3%
   5. PHP, MySQL and UTF-8?.............................:  10  1.5  8.3%
   6. date parsing......................................:   6  1.0  5.5%
   7. What would you like to see in most in a text edito:   5  0.8  4.8%
   8. snmpset with type 'a'.............................:   5  0.8  4.7%
   9. Notice: Undefined index: action error please help.:   5  0.7  4.2%
  10. Startup City 2011 (September 24th,Mumbai).........:   4  0.6  3.5%
 
 Top 10 Original Content Ratings: (Original Bytes) / (Orig. + Quoted Bytes)
 ==========================================================================
   1. Jerry Stuckle.............................................:  37.4%
 
 Top 10 Crossposting Groups: Posts in Group
 ==========================================
 (Articles cross-posted to > 5 groups are not considered in ANY stats.)
   1. alt.support.diabetes.........................................:   1
 
 ==============================
 End of stats for comp.lang.php







-- 

[toc] | [next] | [standalone]

#3199

On Sep 19, 1:21 am, me <m...@privacy.net> wrote:
>  "Caveat: Quantity is not necessarily a measure of Quality"
>  ..........etc

I recently spoke with a guy working for one the search engines, who is
working on automatically determined forum personality profiles. They
want to weight certain posts higher or lower in query returns,
depending on the "personality type."

The idea is this: most forums have dominant participants--those who
post more than others, and those whose posts get responded to by
others.  There are several things to look at:

For this frequent poster, what is his or her ratio between original
thread starts, created by that frequent user, to posts that follow up
to posts created by others.  Frequent posters who seldom offer
original thread head contributions (who only respond to headings
created by others) are flagged as possibly suspicious. If the ratio of
phrase patterns in those follow up posts contain an above average
ratio of well-known angry or aggressive phrases, that personality's
influence is discounted.

If that frequent poster has a high ratio of original thread head
offerings, he or she is flagged as a possibly important and positive
influence.  If others follow up to those original offerings, with
posts not containing a higher than normal ratio of well-known angry or
aggressive phrase groups, then that frequent poster is confirmed as a
personality of importance, whose posts are moved up higher in keyword
search result groups.

[toc] | [prev] | [next] | [standalone]

#3201

pittendrigh wrote:
> If that frequent poster has a high ratio of original thread head
> offerings, he or she is flagged as a possibly important and positive
> influence.  

Or a troll.

> If others follow up to those original offerings, with posts not
> containing a higher than normal ratio of well-known angry or
> aggressive phrase groups, then that frequent poster is confirmed as a
> personality of importance, whose posts are moved up higher in keyword
> search result groups.

So basically feeding the trolls.

The importance of a very frequent poster cann still be lower then the
occasional posting by one.


houghi
-- 
First we thought the PC was a calculator. Then we found out how to turn
numbers into letters with ASCII and we thought it was a typewriter. Then
we discovered graphics, and we thought it was television. With the World
Wide Web, we've realized it's a brochure. -- Douglas Adams.

[toc] | [prev] | [next] | [standalone]

#3202

On Sep 24, 8:41 am, houghi <hou...@houghi.org.invalid> wrote:
> pittendrigh wrote:

My quick characterization undoubtedly does injustice to the algorithm
that guy is working on.
The attempt is to identify trolls cyber bullies of all kinds, and to
discount their influence in search results.

Trolls (those whose primary motivation is simply to cause trouble)
exist and they are a problem.
There are others whose primary motivation is to aggressively dominate
a group, as if the discussion was somehow their possession.
Some search engine engineers are hoping to identify all such
negatively-associated personalities, and to somehow discount their
importance.

I probably should not have attempted to indicate how all of this is
supposed to work. The important concept is that the search engines are
trying to do it all--to favor some personality profiles over others.
That much is interesting.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.php

csiph-web