I named this project Interviews Visualized and it was a fun way to combine the art of storytelling, data design, NLP algorithms, and IBM Watson Natural Language Understanding.
I included data storytelling in the traditional format of a newsletter I run for the team I work in IBM.
By extracting additional meaning from the stories I would write, I wanted to provide readers with a second layer of information: all the information they could not see in the text itself.
To give readers the ability to see the patterns existing in the text’s words, I combined NLP techniques, IBM Watson NLU and data design to visualize the structure and semantics of the words used in the newsletter. In this way, I transformed the data generated as the output of a model into a compelling visual story.
AI driven Stories,
my role in the project:
Design the visual story
Collaborate with data scientists to scale my initial NLP model in a full-fledged engine that I named interviews processor
Following this concept, I came up with a visual way to represent all this information and attached a legend to every visual story to encourage people to take the time to explore the data and go deeper into the details.
The ultimate purpose of this experiment, which kept going for several weeks with several visual data narratives, was to find new ways to make AI and ML technologies more accessible to a broader audience of non-experts and to create new strategies to communicate the content generated by algorithms and models.