- 1) by polling the chat rooms that we have in our databases.
- 2) by pulling data feeds from aggregators.
About online activity statistics
posted by wonderwall on 2014-04-04 14:07:55 GMT
Online activity statistics is a feature that Camroster has and most of our members consider it to be a valuable resource for them. Chat hosts and chat room operators use the statistics to get a picture of their chat activity and see how they have utilized their online time. Frequent chat room users and visitors are able to see the time patterns of online presence of chat rooms and arrange their own schedule.
At this point we have to stress that the activity Statistics should be considered as indicative, they are NOT 100% precise and the reason is the nature of the data availability.
Most of the chat rooms that we monitor are http based, http is a stateless protocol, meaning that the server does not retain any information during each connection making it unsuitable for real time operations. This technical constrain defines the ways that can acquire our raw data and we collect them in two ways:
Data collection and data processing are executed periodically every n seconds, where n is between 3 and 10, as a consequence the real start and end times of a chat session are actually within a time space that is defined by n.
Apart from this peculiarity, a problem that we have to deal with, is that some times the data feeds that we pull from the aggregators are not consistent. Information for active chat rooms may disappear and reappear in an irregular way and we have to make a "guess" here that the chat room was active during a time interval which we have defined to be 45 seconds.
Another problem with the feed aggregators is that some chat rooms do not appear in the feeds when they enter a closed chat session, in this case we poll the chat rooms individually to check if they are active. The polling of those rooms takes place as long as they are online and not included in the feeds.
We hope that this post clarifies the ways that we collect data for the purpose of compiling statistics and clears out any questions you might have.