Watson is inspiring musical creativity with Grammy-winning music producer Alex Da Kid, who used IBM’scognitive services to inspire his new song about heartbreak, “Not Easy”. To help the song top the charts, Watson pulled from his collection of APIs to unlock novel insights and concepts from five years of culture and music data. Watson Distinguished Engineer, Dr. Amini’sdetails the project in herIBM Think blog.
How was Watson trained to understand both culture and music? The creative process was simple. To understand culture, AlchemyLanguage API first uncovered the most pervasive themes in each year. Then, Watson Tone Analyzer read blogs, news articles, and social media to gauge the sentiment around those themes. With this data, Watson arrived at an emotional fingerprint of culture. To understand music, Watson Tone Analyzer read the lyrics of the top 100 songs for each week in the last 5 years. Thus, Watson arrived at an emotional fingerprint of music. To analyze both culture and music, Watson read 2.8 million lines of text. In addition, Alex experimented with Watson Beat, a technology that curated brand new music beats as inspiration for his next song.
This project demonstrates the broader impact cognitive computing has wherever humans and machines interact. Where we once had to manually analyze big data sets to see if there was negative or positive sentiment on a product, AlchemyAPI is able to do so in minutes, first by extracting sentiment, followed by a deeper analysis to detect distinct emotions and broader concepts. Try using AlchemyAPI to analyze your own text here.
Pairing perfectly with AlchemyAPI, Watson Tone Analyzer then provides advanced linguistic analysis of written text. Tone Analyzer detects tones from expressed emotions, social tendencies and writing style. This API can be used for things as simple as receiving feedback on communication style to being incorporated as a ‘digital virtual agent’ to monitor and flag frustrated clients interacting with automated agents.
Much in the same way, Watson APIs are used to propose evidence-based answers in areas like cancer research, pharmaceutical discovery, finance, and environmental and ecological exploration. Most importantly, Watson APIs are easy and simple to set up. Start analyzing your data immediately and building apps quickly.
In the fight against shrinking ad dollars, fake news, ever smaller editorial and news staff and increasingly dominant social media platforms, Oovvuu is using Watson to better connect viewers to the most relevant video and news content, generating valuable new advertising revenues for media and news organizations. Their platform can match up to 1000 videos to relevant news articles in less than a second — a 2400% increase over manual methods.