Watson Studio, Watson ML, Watson OpenScale

 View Only

An Exciting Future for IBM SPSS Modeler

By Ted Fischer posted Wed November 22, 2017 08:31 PM

Last month my colleague Doug Stauber announced a new user interface for IBM SPSS Statistics that is coming with the 50th anniversary of SPSS Statistics product. I am very pleased to announce that IBM continues to invest in the SPSS brand and product roadmap. IBM SPSS Modeler, a favorite tool of so many data scientists and analysts, will also be getting a new attractive and friendly user interface and additional enterprise-ready and user-requested features over the next several months. This next generation of IBM SPSS Modeler will work with IBM Data Science Experience on cloud and IBM Data Science Experience Local – and also allow customers who wish to only use IBM SPSS Modeler visual data science capability to continue to do so as well. The new interface is also planned for the newly announced IBM SPSS Modeler Subscription.

In 2017 IBM SPSS Modeler added new features to support the needs of non-coder users such as open source algorithms like Spark and Python available from the GUI interface and seamless integration with CPLEX allowing optimization capabilities straight from SPSS Modeler. It improved big data integration, added new functionalities for text analysis and made Modeler more accessible than ever with a monthly subscription licensing starting $199 a month.

In 2018, IBM will keep its effort to make IBM SPSS Modeler the data science tool for non-coders, but also totally integrated with IBM Data Science Experience products allowing users with different needs, preferences and skills to work in collaboration. IBM SPSS Modeler will be even more accessible, user friendly and attractive to it’s current and future users. IBM is committed to bringing Data Science to All, enabling both the non-coder and the coder to collaborate within the same platform.

Modern, new interface

The new interface will be accessible via a web browser though still also available on the desktop and will incorporate contemporary design principles. The basis of this new user interface is already available today in IBM Data Science Experience as an open beta feature. It is currently labeled as “Machine Learning Flows” (although the label will soon change to “SPSS Modeler”). We are incorporating user feedback as we change this interface – please try it out and feel free to add your comments below.

Improved Visualization

In addition, we plan to build on the new visualization already exposed in Modeler 18.1.1 as a beta feature – with the end result that the data scientist can confidently create great visualizations that can be presented to anyone.

Visual Data Preparation

We also intend to add a visual data preparation capability that draws upon the principles exposed in the Data Refiner available in Watson Data Platform as a beta feature. In Data Refiner, a non-coder can prepare or transform data using a few clicks with the results immediately visible. Significantly the non-coder can create a defined flow of operations than then can be executed again and again.

Improved Server Installation

Finally, we plan to ensure that the server installation and configuration experience for IBM SPSS Modeler’s deployment capabilities will be much simpler – while ensuring valuable features such as deployment and pushback to Hadoop continue to be market leaders.

Vision of an Integrated but Modular Platform

IBM is working toward a vision of offering the most complete enterprise Data Science platform, allowing access to Data Science tools such as IBM SPSS Modeler, Data Science Experience, Watson Machine Learning and Decision Optimization but giving customers the choice of which capabilities they want. Our goal is to have these capabilities available on public cloud, on private cloud/on-premises and where applicable on desktop. For the enterprise, our goal is that the platform will provide a comprehensive set of deployment, scalability, governance and security features.

This new interface will be valuable in many ways. For the existing IBM SPSS Modeler customer, we are doubling down on our commitment to deliver easy to use, modern visual modeling software. For the data science coder who needs to work with non-coders, the addition of one of the leading software packages in visual data science to IBM Data Science Experience means that the coder can now much more easily collaborate with the non-coder in the same environment. This is one of the most requested capabilities we hear from our customers: how can I empower more data scientists in my organization, and how can working together be faster, easier.

Please subscribe to this blog if you would like to hear more about the IBM SPSS Modeler. Also, login to datascience.ibm.com today to try out the new interface.

1 comment



Thu November 23, 2017 04:11 AM

This is great !! Really exiting about this new experience 😊