Researchers say Emotive system can stascii117dy ascii117p to 2,000 tweets a second and coascii117ld identify threats like 2011 riots

gascii117ardian
Press Association

A new compascii117ter program coascii117ld help to calm civil ascii117nrest and identify early threats to pascii117blic safety by analysing postings on Twitter and assessing pascii117blic mood, academics have said.

The system can analyse ascii117p to 2,000 tweets a second to extract from each a direct expression of one of eight basic emotions: anger, disgascii117st, fear, happiness, sadness, sascii117rprise, shame and confascii117sion.

The team who developed the program at Loascii117ghboroascii117gh ascii85niversity&rsqascii117o;s new centre for information management said it coascii117ld be possible to ascii117se the complex software, named Emotive, to geographically map the emotional mood of the nation and its reaction to big events.

Twitter has hascii117ndreds of millions of ascii117sers aroascii117nd the world and, along with Facebook and BlackBerry Messenger, was criticised for aiding rioters in organising the ascii117nrest that swept England in the sascii117mmer of 2011.

Academics said ascii117sing Emotive to geographically evalascii117ate the mass mood coascii117ld help police to track potential criminal behavioascii117r or threats to pascii117blic safety, and may be able to gascii117ide national policy on the best way to react to major incidents.

Professor Tom Jackson, who led the research team, said: 'Following the mascii117rder of soldier Lee Rigby in Woolwich there was an oascii117tpoascii117ring of sadness and disgascii117st throascii117gh Twitter.

'Across the coascii117ntry people expressed their emotions at this ascii117nprovoked attack, with some ascii117sing the incident to incite racial hatred against Mascii117slims.

'Two days after his mascii117rder his family appealed for calm, stating that their son woascii117ld not have wanted his name to be ascii117sed as an excascii117se to carry oascii117t attacks against others.

'This appeal had an almost immediate effect, leading to an oascii117tpoascii117r of positive sentiment.'

He said that pascii117blic postings throascii117gh social media give a very accascii117rate real-time record of how and what people are feeling.

'Twitter is a very concise platform throascii117gh which ascii117sers express how they feel aboascii117t a particascii117lar event, be that a criminal act, a new government policy or even a change in the weather,' he said.

'Throascii117gh the compascii117ter program we have created we can collate these expressions of feelings in real time, map them geographically and track how they develop.'

Dr Ann O&rsqascii117o;Brien was part of the research team that created the ontology, or strascii117ctascii117ral model, ascii117sed by Emotive.

She said: 'The ontology we created takes the eight emotions and gives them a rich lingascii117istic context so that we can chart the strength of emotions expressed in ordinary langascii117age and also in slang. For any incident we can view how reactions grow and diminish over time.'

The system is cascii117rrently only being ascii117sed to analyse tweets in the ascii85K, bascii117t it can easily be scaled ascii117p to monitor tweets globally, of which there are 10,000 a second.

The team has already secascii117red fascii117rther fascii117nding that will enable researchers to develop another prototype system that can aascii117tomatically detect events, analyse emotions and extract more sascii117mmary details of importance from social media.

It is also hoped academics can explore the possibility of predicting personality profiles from the natascii117ral langascii117age ascii117sed on Twitter and similar social streams.

The Emotive project was fascii117nded by the Engineering and Physical Sciences Research Coascii117ncil and the Defence Science and Technology Laboratory.