Sentiment Analysis Project

Sentiment Analysis Project allows you to discover patterns and relationships that would otherwise remain hidden.

Importantly our app uses real-time data together with historical sources to give a truly comprehensive view of the selected themes.

The app reveals fascinating insight through comparison. Does negative social sentiment around one company raise the stock price of its main competitor? How do the antics of a pop band affect the fortunes of their record label?

Sentiment Analysis Project was designed to be part of the 2013 London Design Festival (LDF). It featured as a large installation in a media and arts space Tent London in Shoreditch, London’s digital media hub.

As a part of our public relations campaign, we used the occasion to talk to a broader audience interested in design, highlight our expertise, promote our product and attract talent.

At this stage the app is experimental. In the future it will be integrated in our core application Eikon, an advanced platform for financial and risk professionals. More information can be found on the website thomsonreuterseikon.com

Our main success was bringing data visualisation to a broader audience. The app showed people the importance and beauty of interpreting and visualising data – and how good design can bring meaning to something that is otherwise inexplicable.

Of thousands of visitors, our team spoke to hundreds of people directly. We got in touch with a diverse group of people enjoyed using the app, discovering more about data visualisation and our work.

As part of the project we were invited to give a related key note talk at the Victoria and Albert Museum in London.

We’re an expert group of researchers, strategists, creatives, user-experience specialists, designers and software engineers. Every team member is multi-disciplinary, providing an additional depth of skills.

We used a blend of novel HTML 5 technologies. We used Node.js to display seamless real-time tweets and news and to enable interaction between screens. Google’s AngularJS to give structure to our front-end code. HTML5 Canvas for data visualisations. And MongoDB for rapid data storage.

Significantly, we also wrote a bespoke algorithm, to analyse data and grade sentiment within it. Since we used browser-based technologies, our system is hardware agnostic.

To ensure our message was truly engaging we used a giant five by two metre screen built to our specifications. The screen was connected and communicated to a 40 inch multi-touch surface that allowed visitors to easily interact with the app and the visualised data.