The Pulse. A real-time, data-reactive, living key visual. The Pulse is a dynamic, data-responsive graphic that tracks over 10,000 real-time conversations on Twitter.

This revolutionary tool uses real-time Twitter-data and sentiment-artificial intelligence to translate social conversations around women in leadership into a living brand asset. Introducing: The Pulse. Developed for Women Political Leaders (WPL), a premier political women’s network with influential members from around the world. The Pulse is a dynamic, data-responsive graphic that reacts to six key topics: Power, Success, Community, Diversity, Change and Leadership – all representing WPL’s key values.

Client

Women and Political Leaders (WPL)

Agency

Landor Associates, London

My roles

Concept & Design Development of interactive branding system Development of sentiment analysis solution Development of Twitter-data acquisition Development of data storage and backend-/frontend implementation

Formats

Marketing Collateral / Keyvisuals Interactive Experience Large-scale projection

The Pulse was used to generate all key visuals for the Women Leader’s Global Forum event in Reykjavik, Iceland. The event brought together international political and global leaders.

The Pulse is a dynamic, data-responsive graphic that reacts to six key topics: Power, Success, Community, Diversity, Change and Leadership – all representing WPL’s key values.

When viewed over time, the shifting currents of the grassroots, global debate around women in power are revealed. A snapshot of the conversation and the individual Tweets comprising it can be taken by the viewer at any moment.

www.seethepulse.com

The Pulse’s goal is to be a stimulant for action: To listen, respond and learn – according to global conversation. And to visually remind us of the progress.

The Pulse assigns a unique form to each topic. As topics trend, those forms draw the user’s attention towards specific issues by growing in size. The positive or negative sentiment that drive the ongoing discussions are reflected by a change of color and allow for a quick interpretation of current attitudes.

This approach of applied Generative Design shows how humans and computers co-create new, more democratic stories in radically different ways. As algorithms carry out the duplication of creative models, new forms of communication emerge that are at eye-level with the audience and allow people to shape and influence future narratives.

This approach of applied Generative Design shows how humans and computers can co-author new, more democratic stories.

As algorithms carry out the duplication of creative models, new forms of communication emerge that are at eye-level with the audience and allow people to shape and influence future narratives.

1.

The conversation starts at the Women Leaders Global Forum in Iceland on Nov 26. Looking at 6 key drivers (Success, Community, Power, Diversity, Change and Leadership), every woman in the room will be connected to the world’s collective voice.

2.

The Pulse is listening, in real-time, to the online dialogue that is happening around the world on various social channels. In a heartbeat, it assesses how positive or negative, large or small, the conversation is and constantly morphs its shape to match it.

3.

It becomes a living entity. The more the world changes and adapts, the more data passes through, and it is all captured and understood by The Pulse. It is an algorithmic ear to the ground in a noisy world. The conversations these women want to hear.

The Pulse leverages modern technologies to deliver on its promise of an interactive, real-time, 360° experience.

Its user-facing components were developed using WebGL (powered by three.js), a high-performance graphics-pipeline, so that it can be experienced on all devices, ranging from desktop-class computers to mobile handhelds.

The backend-portion is comprised of a custom, dedicated node.js express-server which asynchronously scrapes relevant Tweets using the open Twitter-API and then hands those off to a TensorFlow-neural net trained on more than 1.4 million classified Tweets. For later analysis, anonymized data of each day is stored in a secure database so that we will be able to look at the state of conversations on days of special interest.

↑ View of the backend’s real-time analysis of incoming Tweets. Tweets are segmented into six key topics, analyzed by the Tensorflow-AI and then immediately served to the frontend.

The backend analyzes Tweets continuously and asynchronously in real-time while using a TensorFlow artificial intelligence (AI) neural network trained on 1.4 million Tweets. The recurrent neural network (RNN) is a form of recursive artificial neural network best suited for natural language processing (NLP).

↑ This radar-chart was used to get insights and first impressions of how sentiment-analysis and Tweet-frequencies aligned. It was later used to fine-tune the visual output.

The results are served via a custom node.js express server which delivers real-time JSON-data including concatenated data-streams as well as detailed information for each Tweet, it’s sentiment score and life-time data. This data is picked up by a custom WebGL-application, developed using three.js, which allows the user to experience the data seamlessly in a 3D-environment. The result is a living, breathing, growing, data-reactive, interactive, 3D brand experience on the Web.

This illustration shows my initial concept of the project and is an overview of the logic and flow of data of The Pulse. The flow is data is collected by the backend, analyzed by the sentiment component and then both served to the frontend for immediate visualization and returned to the “historic” portion of the backend for later use.