Is the future of photo libraries with auto-tagging and Watson?

Tagging digital content is time consuming for a person and very hard for a computer. Who remembers when Flickr’s auto-tagging was called out? (read more here) So needless to say its not a trivial task, unless you are Watson…

I am fascinated by this technology. To be able to actually pull apart an image and apply tags that describe what is in the image is amazing. If you think about the applications here, they are endless. I was thinking about how I constantly look for images on my Mac. I have over 10,000 photos on my mac and the only way I can find them is using the map if I recall where I took it or I remember the time frame and find it based on date. Imagine if you could do a search based on tags!

Since Watson Content Hub(WCH) does the auto-tagging for me I decided to whip up a short video showing the power of this amazing feature using WCH. Remember, you could always build your own application like this article explains using OpenWhisk and the API’s on BlueMix. To get you started, below is a screen shot of the Watson API’s on BlueMix:

Now for the video. Like I mentioned, I was really just playing around with various photos in my collection to see how well Watson did at tagging. All in all, I am very impressed…

Where do you think this technology is going? Will you expect your photo albums to have this ability in the future?

 

Using Watson Content Hub as a central repository and tag generator

Want to learn how to integrate Watson into your applications? Well, it’s as easy as clicking on this link and getting started. From image recognition to financial services API’s there is an entire catalog of Watson API’s right at your finger tips. I personally have been playing with the Watson Content Hub API’s and they are very cool. A project I worked on used the Watson Content Hub(WCH) API’s with WebSphere Commerce where we used WCH as not only a central repository but also extracted the tags Watson generated for product images and pushed them into the product record in WebSphere Commerce. This enhanced the searching for products by a customer greatly. Watson generated key words we didn’t think of at first, making search and type-ahead much better.

Here is a quick video where we used a City Cool product called Minimals Moped. We pushed the images to WCH, Watson generated new tags, then we used those tags in the products keyword field. The search indexer did the rest and viola, the search experience was greatly enhanced to have a broader set of key words per product.

Watson drives IBM Digital Experience

If you missed it, at the IBM Connect conference, Chris and Rob demonstrated how to build a page using Watson to assist in suggesting the content for the page with a single click.  Very impressive and as Chris said “that is awesome”.  Totally agree!

In the video below (which is much longer than the snippet I provide here), Rob shows the Cognitive Assist feature to help in constructing a new landing page for their campaign.

Cognitive Assistant

Take a look at the video to see how cool this new feature is.

The future of Marketing with Watson

 

If you missed the demonstration at Amplify with Watson and the IBM Marketing platform then you missed a really impressive demonstration.

Marketplaces, Cognitive, and Dynamic Pricing dominated NRF16

This year I did not man a booth or guide people around the IBM pedestals but instead I decided to actually attend the National Retail Federation (NRF) as an attendee. I went to a lot of sessions, visited a lot of booths, and observed our very own IBM booths. As I walked the grounds and even waited in line for the free lunch box I was impressed with the dialog of others around me. Over and over I heard many talking about Watson and cognitive computing which then lead into other discussions about IBM and more specifically dynamic pricing.

The IBM booth was crowded pretty much every day, the turn out and excitement around IBM Commerce seemed to be at an all time high – I could barely walk through the booth!

ibm_nrf_cognitive

 

This was the first year I witnessed a change in the dialog from the usual feature/function discussion of buy online, pickup in store, cart abandonment, catalog management, etc to features that will differentiate a brand in the market. More talk about consolidating brick and mortar and the eCommerce channels. The ugly truth that most companies still have separate teams supporting the same functions on the different channels. This then lead into many discussions around pricing. The theme this year at the IBM booth was dynamic pricing and cognitive (machine learning).

If you are interested in learning more about how this works you might want to check out the IBM Whitepaper “Attracting and retaining customers with insights-driven dynamic pricing“.

“More sophisticated retailers are not just reacting but instead proactively testing various pricing strategies to see what effect they have on their customers; they are sensing and responding,” – link

The paper also goes into discussing the challenges of consistent pricing across channels. The power comes when you begin to mix cognitive learning with dynamic pricing:

“As channels blur and retailers have multiple touch points with consumers, price coordination becomes essential….The holy grail of dynamic pricing is achieved through the application of cognitive computing, a self-learning environment that “Understands, Reasons, and Learns” from inputs to intuitively determine the best prices and promotions for customers in context.” – link

Lastly, the conversations around marketplaces were also very prevalent this year. If you are not familiar with marketplaces think Amazon. Manufacturers are looking to revamp their B2B networks with new user interfaces and shopping experiences that outperform the Amazons of the world. Moving from the traditional green screen ordering system to a friendlier online shopping experience like a Staples.com.

One example I heard was a manufacturer makes a widget which costs $150. It is of high quality and has a life expectancy of many years. However, knock-offs that look identical (literally almost the exact same picture) cost $50 on Amazon marketplaces. The $50 product is manufactured in China and is really a much lower quality with a shorter life expectancy. The problem is when you search on Amazon for the product you essentially see what appears to be the same exact product but one is $100 more – so which one do you think gets sold? This problem can be addressed by using a solution like the Mirakl Marketplace Platform to battle the Amazon problem and also give your interface for your partners a face lift. I had the opportunity to see the Mirakl demonstration live and I was very impressed with the user interface and the management capabilities it offers. It offers a complete vendor management solution where you can not only bring up a vendor very quickly but also get an holistic view of your products performance across vendors. Click the picture below to learn more about Mirakl.

mirakl

Mirakl Marketplace Platform

Now, imagine a platform where cognitive dynamic pricing and marketplaces all work together. A manufacturer can then beat out the Amazon marketplaces by controlling their own marketplace and also get pricing insight through dynamic pricing!