An investigation has revealed that 30,000 Twitter accounts have been set up by IS sympathisers following the murder of James Foley.
In the wake of Mr Foley's murder, video footage of his brutal beheading was circulated via social media. Twitter executives announced they would suspend accounts that posted graphic imagery or "calls to violent actions."
In the investigation conducted by Sky News, it was found that 60,000 accounts expressing pro-jihadi views have been set up in total since May, the research found.
27,000 of these were established after August 20 - the day after Islamic State posted the video of the beheading of the US journalist.
In the 24 hours following the publication on Tuesday night of a video showing the murder of Steven Sotloff, 10% of all references to the video on the microblogging site were found to be positive.
One of the most prolific IS tweeters via Skype, whose accounts have twice been removed from Twitter said he is still tweeting and is proud of his contribution to the Islamic State.
The 19-year-old from Western Europe said: "The reason I started tweeting was to support the mujahedeen by activism."
The analysis suggests that social media suggests sites are struggling to police postings effectively. It was found that as soon as an account is closed down, it is replaced by another almost instantly using a change of name.
After an account was shut down following the publication of a death threat made to a journalist.
One account identified by the researchers posted a death threat to a journalist, was shut down and within days an account bearing a very similar name was posting similar material.
Another account on which explicit pictures of dead bodies was posted was suspended, only to re-emerge under another identity using a similar name.
The new accounts are then promoted by other IS-related Twitter accounts.
As part of the research, analysts defined a pro-IS account as one talking about positively about IS and about violence.
Staffan Truvé, the chief technology officer of Recorded Future, who carried out the social media analysis said: "Clearly, the current manual process where Twitter shuts down accounts that are reported cannot keep up with this tactic.
"We believe automated approaches that use techniques similar to those we have presented here sentiment and network link analysis can be used to successfully block improper content that violates the Twitter rules."