New phone app protects people’s moods

Computer scientists have developed the world’s first mobile phone app which automatically colour codes messages so people know before reading them if they’re likely to make you feel good or bad.

The development, for Android phones, could mean the end of people being surprised by an angry or hostile message, whether it’s from Twitter, Facebook or text.

It would also allow smart phone users to prepare for bad news and allocate time to receive it.

Master’s student Lorraine Chambers and her supervisor, senior lecturer Mohamed Gaber, both at the University of Portsmouth’s School of Computing, will present their breakthrough at a conference in Spain in September.

Dr Gaber said: “We are increasingly sending and receiving information via messages on mobile phones. The rate of growth in this area has never been witnessed – everything from Twitter streams and Facebook messages to direct text messages are coming straight at us all the time on our handheld devices.

“This information has an immense power, whether we are reading a worrying social media news story or a warning email from our manager, messages can upset mood and increase stress level, just as good news and encouraging emails can cheer you up.

“The ultimate objective of this application is to make the user aware of the negative contents they receive so they are able to manage their stress in the best possible way. For example, if most of what is received from social media websites by a user on a particular day was negative, it is important that the user attempts to take an action in order to not get stressed, especially if this may affect the individual’s performance at work and/or their behaviour at home.”

The app works by automatically colour coding incoming messages as green for positive, red for negative and blue for neutral so a user can see before opening any message whether it is likely to be worrying or encouraging.

The Portsmouth researchers were inspired to research and develop the app after a visit by their colleague Mykola Pechenizkiy at the Eindhoven University of Technology, in the Netherlands, who had developed a similar capability for emails on desktop computers, together with his Masters student Erik Tromp. Mykola and Eric have worked with the Portsmouth researchers, Lorraine and Mohamed using state-of-the-art technology for sentiment analysis to classify ‘on the fly’ any textual input received on the user’s handheld device.

The researchers tested the technology on a range of Android mobile phones and find it works faultlessly no matter what each phone’s computational power and memory were. The researchers are working on ways to make it freely accessible via Android Marketplace.

If there’s sufficient demand, it will be made available to users of iPhones and iPads.

The results of the project are reported in a research paper that has been accepted for presentation at 16th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, to be held in San Sebastian, Spain.