Last month, we started looking at how you can work with Watson. We began with one of the Watson “products” that IBM offers, the Virtual Agent. This month, we will look at the other product options in Watson’s stable.
IBM offers a number of products, off-the-shelf components that can be configured to work with your business and will provide benefit without your having to learn a new language or really do anything technical.
We started by looking at the Watson Virtual Agent (VA), which offers a chatbot that can be used to answer customer questions and that learns as it goes along. But VA is not the only thing IBM is offering, and this month we’ll see what else is out there just waiting to be used.
To refresh your memory a bit on how to get to these products, just go to www.ibm.com/watson and click on Explore Watson Products. Then page down.
Explorer is a data retrieval and analysis tool. It picks up both structured and unstructured data from within your company’s system and analyzes it for relationships and trends.
The data retrieval part of this is a big deal. Current statistics (they are IBM’s, so they must be true) say that data analysis people spend 80 percent of their time finding and structuring data so that it can be consumed by an analytic tool. Since Watson does this automatically, for both structured and unstructured content, it frees up the analyst to analyze.
In addition to efficient indexing mechanisms that speed the analysis, the ability of Watson to work with natural language makes it possible for it to mine data that otherwise would have to be carefully prepared. And the resultant databases can be stored either on your machine or in the cloud.
The IBM website does not quote a price for this service, but you are invited to contact IBM for pricing information, and you can’t get much more customer-friendly than that.
Watson Analytics is also a tool for analyzing your data.
In its most basic format, Analytics runs in the cloud, and it’s free. I know the first page of the link above says $30, but that’s for the Analytics Plus. The basic version is free. Just sign up on the IBM site, and you can upload data from spreadsheets and limited analytics.
You upload the data set you want analyzed (you get 1MB of free storage as part of the base version), which to me means that it’s probably looking for structured data. Then, there’s a set of predetermined questions that you can ask Watson to analyze over this data set. Plus, you can ask any question you want in your own words. Naturally, there are dashboards and drilldowns and all the other cute things that make data analysis so sexy. Various predefined templates are ready for you to use, and they make it easy to put your results in a format slick enough for any meeting. And you don’t need to know a single thing about technology or even about data analysis.
There are a couple of other flavors to it, if you’re interested. The first is Analytics Plus. This you have to pay for, but it’s only $30 per month per user, and that’s not bad at all. The Plus version gives you 2GB of storage and allows you to bring in spreadsheets with up to 1 million rows and 256 columns. The other thing that it gives you is the ability to pull in data from Twitter and include that in your analysis.
If that isn’t enough, there’s Analytics Professional. With this, you get 100GB and room for spreadsheets with up to a million rows and 500 columns. You get twice as much Twitter data (didn’t realize it was limited in the Plus version) and the ability to connect to a variety of cloud data sources. Two or more users are able to collaborate and share and view data even though each has his or her own dedicated space.
But I can take one look at you, and I know you want more. So there’s Watson Analytics for Social Media, which lets you connect to a wide range of boards, review sites, and social networks. How cool is that! And actually, for many companies, that’s very important. I just spent some time with a major regional retailer, and they tried to keep very close track of what was happening on the social media scene. It’s not an easy job because social media is unstructured content. Sometimes very unstructured, but Watson makes it easy.
Bottom line, Watson Analytics is a relatively inexpensive way to get your feet wet in the data analysis world without having to know a single thing. Can’t beat that.
Starting at $150 per month per user, the Studio is designed for developers and subject matter experts. Together, they can create annotator components, and I don’t have to tell you how cool that is.
In short, the Studio allows you to teach Watson about the nuances of your specific industry by using annotations in the input material you give it. While this is a higher-level process (more structured) than just screwing around with analysis, it does not require any coding or special technical ability. At the same time, in addition to using dictionaries, annotated documents, and other resources, you can use the Alchemy language (we will be talking more about this when we look at the APIs you can create), so it can get fairly hardcore. I believe Studio works hand in hand with Explorer, but it may play nicely with Analytics too; I’m just not sure about that and the form that you have to fill out to “talk to an expert” is pretty long, so I didn’t bother.
You Want More?
And that’s them: the products (they’re actually more like services in my mind) that IBM has built over Watson: Watson Virtual Agent, Watson Explorer, Watson Analytics, and Watson Knowledge Studio. Each is available on a subscription basis, and you don’t really have to know a thing in order to use them.
But products are not all that IBM offers through Watson. Remember APIs? Well, those are coming up next, so hold that thought.