According to Calaus, the prediction of the behavior of a crowd is possible through the identification of a crucial "indication" in the midst of "noise".
The idea of using the social media, όπως το Twitter, ως μέσου πρόβλεψης μελλοντικών εξελίξεων είναι παλιά σχεδόν όσο και τα ίδια τα κοινωνικά δίκτυα. Μέσα στα τελευταία χρόνια πολλοί έχουν κατά καιρούς ισχυριστεί ότι είναι δυνατή η πρόβλεψη εκλογικών αποτελεσμάτων, πορειών productof on the market, successes (or not) of films, etc., through analysis of data of social media.
One of these is Nathan Calous, MIT, who has developed a method of predicting crowd behavior through Twitter statements and comments. In particular, according to the MIT Technology Review, Calous has analyzed the tweets associated with the 2013 coup in Egypt, and claims that the whole turmoil associated with the incident was clearly predictable from several days earlier.
At a very practical level, it is not difficult to think how the future behavior of a crowd can be mirrored on Twitter, through, for example, reports and updates on mergers and co-ordination through them. Such activity in social networks is a clear indication of concrete mass behavior in the future.
According to Kallus, predicting the behavior of a crowd is possible through the detection of a crucial "cue" in the midst of the "noise". As he points out, this is possible through "catching" tweets, with the aim of identifying references to future events and then analyzing trends (trends) associated with them.
"Replenishment and the realization of a common energy can often be perceived through trends that appear from before in the data," he says.
It is precisely this kind of analysis that deals with a company under the brand name "Recorded Future", which scans 300.000 different "sources" on the Web in seven different languages in order to identify references to future events. Analyzing these data, Calous seeks to anticipate phenomena such as large-scale protests and riots. "We find that the volume of publicly available information has the potential to reveal future multi-faceted actions," he says.
At the first level, a "protest / disorder" is defined as "significant", which receives more coverage than conventional SMEs than usual. This coverage is then analyzed to see when such phenomena occur and the activity on Twitter is being searched for. The aim is to identify "motifs" so that corresponding "sequences" are perceived in other cases. As for Morsi's fall in Egypt, Calus thinks that the riots, which started the sequence of events leading to his removal from power after an army intervention, were predictable via Twitter well before their inception.
Source: naftemporiki.gr
