IBM Watson Client Experience Centre
IBM Watson Client Experience Centre

IBM has delivered a new set of tools for developers looking to take advantage of IBM Watson. The tools are all part of the cloud-based Watson Knowledge Studio (WKS). With WKS, IBM plans to make it easier for developers and domain experts to build new Watson based solutions for analysing unstructured text.

The majority of content held by organisations is considered unstructured. That means it sits outside of databases and includes documents, audio, video and other content. Searching for key words,  phrases or combinations of content is not difficult. Writing applications that applies contextual controls to those searches is.

Who is this aimed at?

The key focus for WKS is about teaching Watson the intricacies of natural language. To enable Watson to create insights around the data it uses, it has to understand the nuances of that data. WKS will allow developers and domain experts to teach Watson what is right and what is wrong.

Importantly, this is a no code solution. This should widen the pool of people available to create new Watson applications. A good example of this is the creation of compliance aware applications.

Inside organisations compliance teams reduce legislation to a set of complex documents. IT and other departments use those documents to configure their systems. Auditors then have to look at those systems to identify compliance breaches. Automating this creates large numbers of key word searches which takes time to analyse. Using WKS, compliance and audit teams could reduce that analysis by developing their own context aware searches. This would reject all instances of harmless use of keyword and flag where words and phrases are a risk.

Four key features of WKS

There are four key features of WKS listed by IBM. These are:

  • Provide an intuitive way to transfer knowledge from humans to computers for text analytics solutions.
  • Administer the full lifecycle of annotator component development within one tool.
  • Create custom annotator components from scratch that extract entities and relations from domain-specific unstructured text.
  • Deploy annotator components from within Watson™ Knowledge Studio to IBM Watson Explorer and IBM Watson Developer Cloud.


One of the major challenges for IBM Watson is how to train it. In May, IBM began the process of training Watson for Cyber Security. This project involves a number of universities training Watson to spot an cyber attack taking place. Customers will not see new solutions immediately as the work to train Watson will take time.

What IBM is doing with WKS is trying to speed up the creation of new Watson solution. Companies may see this as something to use internally rather than build commercial solutions around. It will be interesting to see if the early users are partners, customers or potentially start-ups. The challenge for the latter will be getting large enough sets of data.

So far, IBM has worked with many research groups to create the various headline Watson solutions. Will we now see professional bodies and research teams branch out on their own?


Please enter your comment!
Please enter your name here