[These are preliminary results]
Capturing all the pings to Weblogs.com on the 28th of July showed some interesting temporal patterns (part 1, part 2). In particular, it showed a natural trend of posting times which had a mid-day peak (wrt US time). The next challange is to break down these pings (i.e. blog update events) by location.
The 816k pings came from 503k unique URLs. To simplify I took all the MSN Spaces updates from the total sequence of updates. This yielded 178k pings from 146k URLs. MSN Spaces blogs contain profile information right on the front page. Generally this includes the name, age and location of the author. By crawling these blogs we can capture this information and start to break down the Spaces ping behaviour by some demographic data.
So far I have crawled 30k of the Spaces blogs that pinged weblogs.com on the 28th of June. From these 30k blogs, 25k contained location information. The top 20 countries in this data were:
8197 China
2480 United States
2215 Taiwan
1552 Japan
1201 Brazil
1001 United Kingdom
939 Australia
883 Canada
758 Spain
693 Mexico
474 France
433 Italy
419 Hong Kong SAR
333 Thailand
288 Netherlands
270 Argentina
149 Peru
125 Germany
121 Belgium
117 Singapore
By plotting the pings per hour for a country we can get some understanding of the elements that make up the overall trend seen in Spaces blogging behaviour. Here is an example comparing pings from China with pings from the United States:
Though the results are preliminary (I need to crawl all the Spaces data), this suggests that the trend in posting behaviour is domniated by Chinese bloggers. In addition, it suggests that US Spaces bloggers produced a pretty steady stream of data from around noon through to midnight on the day of study.
Comments