This particular topic was at around 0.85% of all tweets yesterday. What type of algorithm would ignore this traffic, in favor of say around 0.05% (for first topic).
This has some human fingerprints.
Today - some related topics are making it to the trends.
> What type of algorithm would ignore this traffic
Here's are some algorithms that would cause behavior different than what you would expect:
* Take rate of tweets into account
* Take the geography of tweeters into account
* Only allow a topic to trend if it has an odd number of letters
Obviously the real algorithm probably doesn't count the letters in the topic but it could be if it was a bug. It appears your mental model for how something gets into trending topics is:
* Take the topics that have the highest total tweets per day.
That is almost certainly not correct.
> This has some human fingerprints.
You absolutely don't know that. It's pure speculation.
Comments
This particular topic was at around 0.85% of all tweets yesterday. What type of algorithm would ignore this traffic, in favor of say around 0.05% (for first topic).
This has some human fingerprints.
Today - some related topics are making it to the trends.
> What type of algorithm would ignore this traffic
Here's are some algorithms that would cause behavior different than what you would expect:
* Take rate of tweets into account
* Take the geography of tweeters into account
* Only allow a topic to trend if it has an odd number of letters
Obviously the real algorithm probably doesn't count the letters in the topic but it could be if it was a bug. It appears your mental model for how something gets into trending topics is:
* Take the topics that have the highest total tweets per day.
That is almost certainly not correct.
> This has some human fingerprints.
You absolutely don't know that. It's pure speculation.