As we’ve established, I can no longer watch TV without seeing or considering the ways in which data science, artificial intelligence and/or plain old nerdy tech stuff could apply to make a situation better or more practical. Seriously, I can’t stop! I consider the kind of data a scenario would generate, or what data and/or technique would be required to solve a given problem.
My latest such distraction was during an episode of Black-ish.
Centered around the Johnsons, an African-American, upper-middle class, suburban family, Black-ish is a critically acclaimed, conversation starting sitcom about marriage, parenting, in-laws, sibling dynamics, etc., which tackles relevant social issues, namely those related to cultural identity.
Now, while sitcoms such as The Big Bang Theory and Silicon Valley have obvious connections to technology, I wouldn’t have guessed that a family sitcom, albeit a cutting edge one like Black-ish, would inspire me to think up ways to apply cool tech, but it did.
Andre, patriarch of the Johnson family, has quite the closet. I mean, quite the closet. His shoe collection rivals that of my last obsession, the Manolo slinging, Carrie Bradshaw. He is such a sneaker head, in fact, that he has a subscription service that provides him with a fresh new pair of kicks on a regular basis.
Clothing subscription services are popping up everywhere these days, and the most successful ones use data science to determine what to ship out, given reported preferences and ratings on individual items. What if we took that a step further? What if those recommendations could take into account what we already own?
A smart closet!
We live in an instrumented world. How do we instrument for good? Closet good.
The Internet of Things is everywhere. Our phones, appliances, cars, etc., talk to us. My kids’ toys and books now have IoT built in. IBM prolific inventor, Lisa Seacat DeLuca, wrote The Internet of Mysterious Things to introduce my kids (all kids, really) to IoT in a world that we could never have imagined while growing up, except, of course, for the Jetsons making such predictions decades ago.
So, that smart closet. Using NFC or RFID tags, we can track what pieces, colors, fabrics, designs, etc., we tend to wear. How often we wear them. How we style them. That way, our subscription service can pair recommended items with what we already have, filling holes in our wardrobe. The perfect stylist.
Using that sensor data as input, we can further enhance our wardrobe using data science.
With IBM Watson Data Platform, we can integrate usage data, purchase history, clothing attributes, our ratings and mood, online reviews, etc., to identify the most important factors related to satisfaction. And, predict what we might want or need next. Taking that a step further, we can add constraints related to budget, retailers, promotions, seasons, etc., to optimize our ensemble decisions.
Andre’s shoe collection can get even better, if you could imagine, with personalized selections that complement his current wardrobe. Imagine opening a box containing the new Win Like 96 Jordans, a Marc Jacobs bag or Alice + Olivia top with suggestions on which jeans, in your closet, they should be paired with.
This can be accomplished in a coding or non-coding environment.
For non-coders, using SPSS Modeler with CPLEX Optimization, you can pair and prepare local and/or cloud data sources, apply a whole host of modeling techniques, add optimization inputs for decision making and share results for deployment.
Alternatively, if coding is your jam, IBM Data Science Experience can offer you the flexibility to conduct the above steps in an interactive and collaborative environment, using data science tools such as RStudio, Jupyter, Python, Scala, Spark, etc.
Once your predictive model is built, in either environment, it can be scored with Watson Machine Learning.
With the Internet of Things and Data Science, our wardrobes and the world can be greatly enhanced. But wait. What if, in addition to personalized recommendations, we could commission pieces designed specifically for us?
Watson entered the fashion world with a cognitive dress at the 2016 Met gala. Training on 200 images of dresses, designed by fashion house Marchesa, Watson designed the piece, selecting materials and color based on rankings and popularity. Further, using Watson to analyze social media feedback, the dress was brought to life in real time.
While we might not be there quite yet, we’re getting closer and closer to our closets dressing us, circa 1960s Jetsons.
To learn more about how to get started, let’s connect.