Key features include Natural Language Querying, integration with Watson Knowledge Studio, new connectors, and usability and performance enhancements

We’re excited to announce that IBM Watson Explorer v11.0.1 is now available and introduces a whole set of noteworthy new capabilities and enhancements. Watson™ Explorer v11.0.1 expands the key cognitive search capabilities and enables users to:

  • Search naturally using everyday language with Natural Language Query module.
  • Improve search results without manual oversight using unsupervised machine learning algorithms
  • Annotate and enrich data with domain models that can be rapidly developed on the cloud using IBM Watson Knowledge Studio.
  • Receive proactive notifications for saved search queries whenever the underlying dataset changes.
  • Measure search and usage of unified information applications with out-of-the-box reports.
  • Connect to new and updated data repositories such as IBM WebSphere Portal Server and IBM Web Content Manager
  • Productivity and security enhancements.

Watson Explorer + Watson Knowledge Studio

An integration with IBM’s Watson Knowledge Studio (a new offering, pending general availability (GA) to be released on June 27, 2016) allows users to leverage state-of-the-art, unsupervised machine learning techniques for rapid development of custom annotators. Users can then use these annotators in Watson Explorer. This new integrated approach provides significant improvements in time-to-value, and enables developers and subject matter experts to use their own expertise to discover insights more quickly and easily than ever before.

Better Natural Language Querying Capabilities

With regards to the Foundational Components (FC), Watson Explorer introduces Natural Language Querying (NLQ) – a natural language and machine learning-based approach for performing information queries. The approach is radically different from the traditional keyword and semantic searches. It is easier to use and allows for imprecise questions, while the quality of the results increases by an impressive 40 percent. *

The NLQ feature is implemented as a standalone service that integrates with Watson Explorer’s FC, and is supported by ontolection trainer, a new tool for unsupervised machine learning.

Enhancements to the Watson Explorer Application Builder

The Watson Explorer Application Builder features numerous enhancements including:
– Notification improvements
– CSS/JS for search results preview
– A table widget accepting arrays, new search settings and more
– Extended reporting capability (recording user interactions with the app for usage statistics and displaying these statistics in a widget)
– A new chart widget that makes query visualization a breeze, using pre-configured templates.

New connectors

Two new FC connectors are introduced in this version – a refreshed Web Content Management connector supporting the latest versions of IBM Web Content Manager, and a brand-new WebSphere Portal Server connector.

Improvements in the Analytical Components

On the Analytical Components (AC) side Watson Explorer v11.0.1.0 introduces several usability, administration, and security improvements. A new user role allows users who are not system administrators to upload custom annotators (as PEAR files) and add them to collections. Also, admin user activities are logged for auditing. Lastly, Watson Explorer Content Analytics Studio (the well-known rule-based annotator editor) is now offered as a 64-bit application, removing the 4GB RAM limitation of the old 32-bit editor, and increasing the performance and reliability of the tool.

Platform support changes

On the platform side, the Watson Explorer team streamlined the OS portfolio by dropping support for SUSE Linux Enterprise Server (SLES), Windows Server 2008 and Windows 7 32-bit. The latest version of Watson Explorer supports Red Hat Enterprise Lunix (RHEL) on x86-64, POWER Systems and z Systems hardware, and Windows 2012 R2 on x86-64 hardware. See the complete list of operating systems and web browsers here.

Learn more about the new capabilities with our release notes for Foundational and Analytical Components. If there are specific topics you would like to see covered in a blog post, let us know with a comment below.

Webinar 1_21

About Watson Explorer

Watson Explorer combines search and content analytics with unique cognitive computing capabilities to help users find and understand the contextual information they need to work more efficiently and make better, more confident decisions at the point of impact. Watson Explorer is offered in two editions:

Enterprise Edition: Provides enterprise-wide information access and unified information applications capabilities across internal and external data sources and services, as well as the ability to integrate Watson Developer Cloud cognitive services for enhancing, scaling, and augmenting human expertise.

Advanced Edition: Provides all of the same functionality as the Enterprise Edition. In addition, the Advanced Edition includes advanced content analytics capabilities to enable organizations to adapt their information access solution to specific domains and to extract insights from unstructured data to help to identify trends, patterns, and relationships in their data.

The new functionality added in Watson Explorer v11.0.1 is available for users of both editions.

(*) Results achieved using academic test data

9 comments on"Announcing the launch of IBM Watson Explorer v11.0.1"

  1. Name *Kenio Carvalho July 01, 2016

    Natural Language Querying (NLQ) is available only on WEX FC?

  2. Duvier Zuluaga July 05, 2016

    Hi Stefan

    Given that WEX will support both Content Analytics Studio and Knowledge Studio, what are the specific purpose of each tool?? when I am supposed to use CA Studio and when to use Knowledge Studio? There will be specific use cases for each kind of annotators?

    • Stefan Tzanev October 17, 2016

      Hello Duvier,
      CA Studio is a rule-based annotator editor, it runs on-prem, and it ships as part of Watson Explorer Advanced Edition.
      WKS (Watson Knowledge Studio) is our next generation annotator editor. It uses machine-learning approach for supervised training of annotators. WKS is a SaaS offering and it is not bundled with WEX. (Access needs to be purchased separately.)
      Typically, you use rule-based annotators when the rules are clear, fewer and high-precision is required. ML (machine-learning) annotators, on the other hand allow for managing complex and ambiguous models, and handling cases that have never be trained. It is a great way to deal with the ambiguity that’s inherent in cognitive systems.
      We currently work on enabling WKS to handle rule-based annotators in addition to Ml annotators. This will enable clients to save their (sometimes huge) investments in rule-based annotators by seeding the ML models in WKS with their existing rule-based annotators from CA Studio. We foresee that this may be a common scenario in the future – start with some rules and then continue with ML training.
      Please post future questions to our public forum at
      Hope that you find the answer helpful.
      Kind regards,

      • Stefan Tzanev October 17, 2016

        One more thing that I missed to mention: The WKS annotators are limited to RHEL OS only, because the SIRE runtime that runs the WKS annotators on the UIMA pipeline (Watson Explorer Analytical Components) was implemented only for RHEL (RedHat Enterprise Linux). The SIRE runtime was developed and is maintained by a separate IBM group.

      • Hi Stefan
        I have a question about supervised and unsupervised training/learning. I ask this question because I have to communicate this to our client. In the announcement you write that WKS is “state-of-the-art, unsupervised machine learning techniques” but in your answer from 17. october you say “WKS (Watson Knowledge Studio) is our next generation annotator editor. It uses machine-learning approach for supervised training of annotators”. Is it supervised or unsupervised.

        Do you have any explicit definitions of supervised and unsupervised training/learning?


        • Stefan Tzanev November 08, 2016

          Hello, Pål.
          I wrongly state in the main text of the blog post that WKS offers “unsupervised machine learning.” My apologies for the mistake. WKS offers supervised machine learning approach to annotation building.
          Unsupervised ML would not require explicit human intervention in the process of building the annotators. Supervised ML, on the other hand, needs a human trainer as part of the process. WKS uses the supervised ML approach.
          We also plan to add support for rule-based annotators in WKS in the very near future. This will allow current users of WEX Ca Studio to protect their investments in rule-based annotators.

  3. Name *anands December 17, 2016

    Hi Stefan – When you mentioned ‘WKS annotators are limited to RHEL OS only’ in the post dated Oct 17, 2016, does this mean WKS Machine learning annotators cannot be invoked on Watson content analytics studio, installed on Windows OS? What is the resolution for the same?- as we have to clearly convey the ASK to customers, in terms of integrating WKS annotator with WEX content analytics studio..

    • Stefan Tzanev December 18, 2016

      Hello Anand,
      You cannot edit WKS annotators in WEX CA Studio. And this has nothing to do with the level of OS support.
      WKS annotators can be integrated with the WEX AC NLP pipeline. That is, WKS annotators can be used with WEX AC in RHEL environment. But these annotators cannot be ported to WEX CA Studio. You can use your WEX CA Studio rule-based annotators in WKS to seed your work there, but porting annotators in the opposite direction (from WKS to WEX CA Studio) is not supported and we currently don’t have plans to support it.

Join The Discussion

Your email address will not be published. Required fields are marked *