0 / 0
2018 What's New

2018 What's New

Here are the new features for Watson Studio, Watson Machine Learning, and Watson Knowledge Catalog for the year 2018.

Week ending 21 December 2018

Deploy functions in IBM Watson Machine Learning

You can now deploy Python functions in Watson Machine Learning the same way that you can deploy models. Your tools and apps can use the Watson Machine Learning Python client or REST API to send data to your deployed functions the same way that they send data to deployed models.

Week ending 14 December 2018

Watson Studio Desktop is generally available!

Watson Studio Desktop is now Generally Available to try and buy. It's a new edition of Watson Studio for your desktop. Keep your data on your desktop and take advantage of Watson Studio features while you're offline.

Check service status for IBM Cloud

If you're having a problem with one of your services, go to the IBM Cloud Status page. The Status page is the central place to find unplanned incidents, planned maintenance, announcements, and security bulletin notifications about key events that affect the IBM Cloud platform, infrastructure, and major services.

A new tool for building neural networks: NeuNetS (Beta)

The NeuNetS tool in Watson Studio synthesizes a neural network and trains it on your training data without you having to design or build anything by hand.

Week ending 7 December 2018

Decision Optimization in notebooks (Beta)

Decision Optimization is now available in Watson Studio with a seamless integration of the CPLEX solvers in the Python runtime environment. When you create a notebook, choose the Default Python 3.5 XS - Beta of DO environment and the Decision Optimization package is pre-installed.

Annotate and label image files with DefinedCrowd (Beta)

You can now use DefinedCrowd to annotate and label image files.

Migrate projects that use Object Storage OpenStack Swift

IBM Cloud is deprecating Object Storage OpenStack Swift. You must delete your projects that are associated with Object Storage OpenStack Swift or migrate them to Watson projects that are associated with IBM Cloud Object Storage.

You can tell if your project is associated Object Storage OpenStack Swift on the My Projects page by looking at the STORAGE TYPE column.

Week ending 30 November 2018

General availability for streams flows

Streams flows are now generally available.

Choose your IBM Cloud service region

Watson Studio and Watson Knowledge Catalog services are available in multiple IBM Cloud service regions. When you sign up for Watson Studio and Watson Knowledge Catalog, your current region is selected by default. You can now select a different region.

With some offering plans, you can provision services in more than one region. You can now see in which service regions you have Watson Studio and Watson Knowledge Catalog services and switch to services in a different region. Click your avatar and then Change region. You'll see an offering plan name for any instances of Watson Studio and Watson Knowledge Catalog services that you can access across the service regions. You can select a different region so that you can access the projects, catalogs, and data that you saved in that region.

Week ending 23 November 2018

Removal of the Tools menu

The Tools menu is removed from Watson Studio. You access tools within a project. To access the notebook editor, the Modeler canvas, or Data Refinery, create a notebook, modeler flow, or Data Refinery flow asset. Click New asset and then the asset type. To access RStudio, click Launch IDE > RStudio.

Annotate and label image files with Figure Eight (Beta)

You can now use Figure Eight to annotate and label image files.

Week ending 16 November 2018

Project Assets page improvement

The Assets page in your projects now shows only the Data assets category by default. Other asset type tables appear after you add an asset of that type. Click New asset to add new types of assets.

Week ending 9 November 2018

Provision Watson services in the Tokyo (AP-North) region

{: #tokyo} You can now provision Watson Studio, Watson Knowledge Catalog, and Watson Machine Learning in the Tokyo (AP-North) service region in IBM Cloud.

The Tokyo (AP-North) region has the following limitations for these services:

  • If you need a Spark runtime, you must use the Spark environment in Watson Studio for the model builder, modeler flow, and notebook editor tools. The Apache Spark service is not available in the AP-North region.
  • The Real-time Streaming Predictions deployment type is not yet available.
  • Deep learning is not yet available. You can't create deep learning notebooks or deep learning experiments.
  • The Neural Network Modeler is not yet available.
  • The Total Catalog Assets information is not yet available.
  • Profiling of data assets that contain unstructured textual data is not yet available.
  • Multibyte characters are not supported in user-input fields in Data Refinery. Some fields allow multibyte characters but do not display them correctly.
  • Activity Tracker events for services provisioned in the Tokyo region are shown in the Activity Tracker service in the Sydney region.

Adding asset types is easier

You no longer need to enable a tool on the project Settings page to add analytic assets and access their associated tools. All analytic assets are listed on the Add to project menu. By default, the project Assets page shows only the Data assets section. As you add analytic assets, the appropriate sections appear.

SPSS Model operator (Processing and Analytics) for streams flows

In the Properties pane of the SPSS Model operator, all models of all Watson Machine Learning instances are listed. Previously, only models of a selected Watson Machine Learning instance were shown.

You can also see the list of all SPSS models in the Assets tab in the canvas palette. Like other operators and connections, you can drag the model from the palette to the canvas, and integrate it into the streams flow.

Week ending 2 November 2018

Manage authorized users for Watson Studio

You can now change the number of authorized users for your Watson Studio account if you have the Standard or Enterprise plan. Authorized users are project collaborators with the Admin or Editor role. You are billed extra for authorized users when they exceed the number set by your offering plan. Choose Manage > Billing and Usage > Authorized Users.

"Streams Designer" renamed to "streams flow"

"Streams Designer" is renamed to "streams flow" to be more consistent with other types of flows, for example, modeler flows or Data Refinery flows. To create a streams flow, from within a project, click Add to project > Streams flow.

Week ending 26 October 2018

Redesigned home page

The home page you see when you log in to Watson Studio or Watson Knowledge Catalog is redesigned to help you get started quicker. Take a tour! Choose Support > Launch Tour.

RStudio is no longer available in older projects with Object Storage OpenStack Swift

When you launch RStudio within Watson Studio, you must now choose a project. You can't choose a project that uses Object Storage OpenStack Swift. You can choose only projects that use IBM Cloud Object Storage. Check the Storage section on the project Settings page to see which type of object storage the project uses.

SPSS Model operator in streams flows (Processing and Analytic)

Use the SPSS Model operator in a streams flow to run a predictive model that was created in IBM Watson Machine Learning. A predictive model refers to the prepared scoring branch of an SPSS modeler flow in Watson Machine Learning.

Week ending 19 October 2018

"Data flow" renamed to "Data Refinery flow"

Data flows are now called Data Refinery flows to better distinguish them from other kinds of flows in the Watson Studio user interface. For example, modeler flows or machine learning flows.

To create a Data Refinery flow, from the Projects page, go to Add to project > Data Refinery flow.

New population pyramid chart in Data Refinery Visualizations

Population pyramid charts show the frequency distribution of a variable across categories. They are typically used to show changes in demographic data.

To access the charts in Data Refinery, click the Visualizations tab, and then select the columns to visualize.

RStudio in Watson Studio projects in the US South and United Kingdom regions

RStudio is now integrated in IBM Watson Studio projects in the US South and United Kingdom regions and can be launched after you create a project. When you open RStudio, a default RStudio Spark environment runtime is automatically activated. With RStudio integration in projects, you can access and use the data files in the IBM Cloud Object Storage bucket associated with your project in RStudio.

Add access groups for collaborators in catalogs

You can add an IBM Cloud access group to a catalog. All the members of the access group become catalog collaborators with the role that you assign to the access group.

Week ending 12 October 2018

Redact masking method

Redact is a new method of data masking for policies. You can now redact data values in asset columns, which means that data is replaced with Xs to remove information that is, for example, identifying or otherwise sensitive. With redacted data, neither the format of the data nor referential integrity is retained.

Streams Designer (beta)

IBM Streams Designer is back in IBM Watson Studio! We’ve developed quite a few new operators and have added functionality to improve your streaming experience.

MQTT operator (Source, Target)

Use the MQTT operator to stream messages. MQTT is a publish and subscribe messaging transport protocol that is designed to push messages to clients.

You must have your own MQTT broker.

Debug operator (Target)

Use the Debug operator to view the tuples coming from a selected operator. No data is stored.

Cloudant operator (Target)

Use the Cloudant operator to store data as documents in an IBM Cloudant database. Data is stored in JSON format.

Cloud Function operator (Target, Processing and Analytic)

Use the Cloud Function operator to process streaming data and make your server-less functions react to incoming events. You can define multiple parallel workers to increase processing rate.

Python Machine Learning operator (Processing and Analytic)

Use the Python Machine Learning operator to run Python models that do real-time predictions and scoring.

Db2 Warehouse on Cloud operator (Target)

Use this operator to store data in Db2 Warehouse on Cloud.

Email operator (Alerts)

Use this operator to send email to selected recipients. You can embed tuple fields, based on the schema, in the Subject and Body of the email.

Downloading log files

You can add log messages to the Code operators and to the Python Machine Learning operator. Those messages are sent to a log file that you can download from the Notifications pane of the Metrics page while the streams job is running.

Installing Python packages

In addition to the supported and preinstalled packages, your streams flow might need other packages for specific work. For these cases, you can install Python packages that are managed by the pip package management system.

By default, pip installs the latest version of a package, but you can install other versions.

Support for container-based Streaming Analytics instances

Streams Designer now supports only container-based plans of the Streaming Analytics service. VM service plans are not supported.

Performance improvements

The Cloud Object Storage operator and Code operators have been optimized to show improvements in performance.

Deprecated

Geofence example streams flow and the geofence operator are no longer supported.

Week ending 5 October 2018

Annotate data with DefinedCrowd (beta)

You can now use DefinedCrowd to improve the quality of your training data. By involving human judgement offered through DefinedCrowd to interpret text sentiment, you can improve your model input data and raise the model confidence.

Week ending 28 September 2018

Watson Knowledge Catalog Standard plan

The new Watson Knowledge Catalog Standard plan fits between the no-cost Lite plan and the Professional plan, which is an enterprise version with additional capabilities and entitlements. Use the Standard plan while you set up your first catalog and policies.

The Standard plan includes business glossary terms, policies, data lineage, and integration with IBM InfoSphere Information Governance Catalog. The integration with Information Governance Catalog allows clients to seamlessly synchronize metadata between Watson Knowledge Catalog and Information Governance Catalog. The Standard plan includes 500 capacity unit hours and the ability to purchase more to process data preparation flows and profiling activities.

Changes to the Watson Knowledge Catalog Lite plan

The Watson Knowledge Catalog Lite plan is updated with these changes:

  • You can now create one rule that you can use in policies and five business terms.
  • You can no longer view the lineage of assets in catalogs or projects.
  • You cannot make new connections to Dropbox, Tableau, Db2 for z/OS, and Looker data sources. Your existing connections to these data sources remain.
  • The number of assets and collaborators in your catalog is now limited to 50. Your existing assets and catalog users remain unchanged. However, if you have more than 50 assets or collaborators, you won’t be able to add more until you reduce the numbers to below 50, or you upgrade your plan.

Data Refinery: Another place to schedule data flows

You can now add a schedule for your data flow from the data flow's Summary page. In the Runs section, click the Schedule tab for the options.

New operation in Data Refinery for summary calculations

Use the new Aggregate GUI operation to apply summary calculations to the values of a column. You can group the results by the values in a different column. Previously, aggregate functions were only available as code operations. The Aggregate operation is under the ORGANIZE category.

Annotate data with Figure Eight (Beta)

You can now use Figure Eight to improve the quality of your training data. By involving human judgement offered through Figure Eight to interpret mood, intention, or tone for example, you can improve your model input data and raise the model confidence.

Provision Watson services in the Germany region

You can now provision Watson Studio, Watson Knowledge Catalog, and Watson Machine Learning in the Germany service region in IBM Cloud.

The Germany region has the following limitations for Watson services:

  • If you need a Spark runtime, you must use the Spark environment in Watson Studio for the model builder, modeler flow, and notebook editor tools. The Apache Spark service is not available in the Germany region.
  • The Batch Prediction and Real-time Streaming Predictions deployment types are not yet available.
  • Deep learning is not yet available. You can't create deep learning notebooks or deep learning experiments.
  • The Neural Network Modeler is not yet available.
  • The Usage Statistics page for catalogs is not yet available.
  • The Total Catalog Assets information is not yet available.
  • Profiling of data assets that contain unstructured textual data is not yet available.

RStudio in Watson Studio projects in the Germany region

RStudio is now integrated in IBM Watson Studio projects in the Germany region and can be launched after you create a project. When you open RStudio, a default RStudio Spark environment runtime is automatically activated. With RStudio integration in projects, you can access and use the data files in the IBM Cloud Object Storage bucket associated with your project in RStudio.

Week ending 21 September 2018

New options on the Manage menu to replace the Admin Console

The administrative tasks that you previous performed on the Admin Console are now on separate pages that you access from the Manage menu.

You must be the owner or an administrator of the IBM Cloud account for your Watson services to perform these tasks:

  • Choose Manage > Storage Delegation to configure IBM Cloud Object Storage to enable non-administrative users to create projects and catalogs and to use your own encryption key.
  • Choose Manage > Environment Runtimes to view and manage the active environment runtimes.
  • Choose Manage > Catalogs to view statistics for all catalogs and assign Watson Knowledge Catalog service administrators.

IBM Message Hub is renamed to Event Streams

As of September 17, 2018, IBM Message Hub is now Event Streams. The name change will not affect how you experience IBM Message Hub; however, it will improve naming consistency across the IBM Cloud catalog of services.

New and improved visualization charts in Data Refinery

Data Refinery introduces new visualization charts with a new user interface that give you more views of your data to better help you explore your data as you refine it. No syntax required.

Enhancements include:

  • Many more charts!: We have 21 charts out of the box and more coming soon. New charts include the 3D chart that displays data in a 3-D coordinate system by drawing each column as a cuboid. You can even rotate the view. The t-SNE chart is useful for embedding high-dimensional data into a space of two or three dimensions, which can then be visualized in a scatter plot.

  • Charts are interactive: Within the same chart, use sliders and settings to view the data in different ways. For example, view a pie chart in a rose or ring format. In the heat map chart, adjust the category order by "As read," "Ascending," or "Descending." Hover over the information to zoom in on the values.

  • Charts are customizable: Select a global color scheme for all the charts.

  • Charts are downloadable: Download a chart as an image with annotated details such as a title and secondary title.

Access the charts in the same way as before. Click the Visualizations tab in Data Refinery, and then select the columns to visualize. The suggested charts for the data are indicated with dots next to the chart type. The chart automatically updates as you refine the data.

New operation in Data Refinery for extracting date or time values

Use the Extract date or time value operation to extract a portion of a date or a timestamp value so that you can use that value for further data refining. The Extract date or time value operation is under the CLEANSE category.

Week ending 14 September 2018

Watson services don't require an organization

You no longer need to specify an organization when you sign up for Watson Studio or Watson Knowledge Catalog and the organization information is removed from the Profile menu and your Profile page.

Support for custom components in AI models

You can now define your own transformers, estimators, functions, operations, classes, and tensors in models you train on IBM Watson Machine Learning for online deployment.

Upload individual images for Visual Recognition

You can now upload individual training images one at a time for use with the Visual Recognition model building tool.

Add service IDs to projects

You can now add service IDs that you created in your IBM Cloud account as collaborators to your projects.

Week ending 07 September 2018

General availability for Spark environments for notebooks

Spark environments for notebooks are now GA. With Spark environments, you can define the hardware and software configurations to start custom Spark clusters on demand. Spark environments can be quickly scaled up or down for resources. This makes them well suited for a variety of use cases from trying out new machine learning algorithms on sample data to running large production workloads on the distributed computation engine.

Publish a dashboard to a catalog

You can now publish a dashboard to a catalog so it can be added into other projects. When you publish the dashboard, you can include a preview of the dashboard so that users can see what the dashboard looks like before adding it into another project.

Watson Studio supports globally provisioned services

You can now associate services with Watson Studio using a combination of resource groups, organizations, and regions. You are no longer restricted to using service instances provisioned in the same region as your Watson Studio instance.

Week ending 31 August 2018

Refresh the preview of connected data

You can now see when the data in the preview was last fetched from the connection and refresh the preview data in projects and catalogs by clicking Refresh.

Week ending 24 August 2018

Db2 Big SQL connector available

Projects and catalogs now support connections to Db2 Big SQL, enabling you to store and retrieve catalog data there.

Spark environments for the model builder and Spark modeler flows (Beta)

Apache Spark environments are available not only for notebooks, which need to run with Spark, but also for the model builder and Spark modeler flows. Instead of using a Spark service in Watson Studio, you can now create Spark environments in which you define the hardware and software configuration of the runtimes that you want your tools to use.

Week ending 17 August 2018

General availability for Natural Language Classifier in Watson Studio

The Natural Language Classifier tooling in Watson Studio is now GA. While maintaining existing API functionality, you can train and test your classifiers within Watson Studio.

Partitioned data assets

You can now create a connected data asset from a set of partitioned data files that are in a single folder in IBM Cloud Object Storage. Partitioned data assets have previews and profiles and can be masked like relational tables. However, you cannot yet shape and cleanse partitioned data assets with the Data Refinery tool.

Week ending 10 August 2018

Looker connector

Projects and catalogs now support connections from the Looker platform.

Tableau connector

Projects and catalogs now support connections to Tableau, enabling you to store and retrieve catalog data there.

Week ending 3 August 2018

Update Natural Language Classifier classifiers trained outside of Watson Studio

You can now update older Natural Language Classifier classifiers by associating the classifiers with a project in Watson Studio and importing the training data.

Lineage for data assets

You can now see the history of the events performed on data assets from files and connected data assets in Watson Studio and Watson Knowledge Catalog in a lineage graph. Only assets that are created on or after 20 July, 2018, have lineage graphs.

Spark environments (Beta)

Apache Spark environments are available in beta. Instead of using a Spark service in Watson Studio, you can use the Spark engine that is available by default for all Watson Studio users.

Week ending 27 July 2018

Microsoft Azure Data Lake Store connector

Projects and catalogs now support connections to Microsoft Azure Data Lake Store, enabling you to store and retrieve catalog data there.

Download the log for a data flow

You can now download the log for each data flow run. Go to the project > Assets tab > Data flows section and click the data flow run. Select View log from the run's menu, and then click Download. The log file name is the name of the data flow appended by the date and time (24-hour system) when the data flow was run. Invalid characters for file names are changed to an underscore (_).

Week ending 20 July 2018

Lineage for Watson Machine Learning models

You can now see the history of the events performed on Watson Machine Learning model assets in Watson Studio, Watson Knowledge Catalog, and Watson Machine Learning in a lineage graph. Only assets that are created on or after 20 July, 2018, have lineage graphs.

Automatically synchronize assets with Information Governance Catalog

You can configure automatic synchronization of data assets between an Information Governance Catalog and a Watson Knowledge Catalog catalog. After the initial synchronization, subsequent changes to data assets in either catalog are propagated to the other catalog. You must configure an Information Governance Catalog - Watson Knowledge Catalog Connector to communicate between the two catalogs.

New "Conditional replace" operation in Data Refinery

Use the Conditional replace operation to replace the values in a column based on conditions.

Additional support for date and timestamp data types in Data Refinery

  • GUI operations: The Replace missing values operation and the Calculate operation can now be used on date and timestamp columns.
  • Coding operations: Many of the commands are enabled for date and time data types. You can use the command-line help and the customized templates to ensure that your command syntax is supported.

Week ending 13 July 2018

Unicode characters in asset names and descriptions

You can now use any Unicode character, except control characters, in asset names and descriptions in projects and catalogs.

Now in the UK South service region: deep learning

Deep learning features make it possible to train complex neural networks on large data sets, using GPUs and distributed training.

Tools now available in Watson Studio web interface:

  • Neural network palette in the flow editor
  • Experiment builder

Actions now available with the command line interface and API:

  • Running one or more training runs
  • Running experiments

Catalog admins can change the owners of public assets

A catalog administrator no longer needs to be a member of a public asset to change the owner of the asset to another asset member.

View the name of who initiated the data flow in Data Refinery

In the Runs section of the data flow details page, the History tab displays detailed information about each data flow. Previously, it displayed the email address of the person who initiated each flow. Now it displays the name of the person who initiated each flow.

Filter multiple columns in Data Refinery

You can now specify conditions for multiple columns in a Filter operation. Previously, you could only select one column. Also within the Filter operation, you can now use the operators “Is equal to” and “Is not equal to” on the values in a column.

Change the source of a data flow

You can now change the source of a saved data flow. This enhancement means that you can run the same data flow but with a different source data asset. The new data set must have have a compatible schema to the original data set (for example, column names, number of columns, and data types). Go to the project > Assets tab > Data flows section and click the data flow. In the data flow summary page, click the "Change the source data asset" icon to select a different data source.

Week ending 29 June 2018

Natural language classification (Beta)

You can now try the beta version of a new, graphical tool in Watson Studio for building natural language classifiers.

Building a classifier is fast and easy: Upload training text examples in a .csv file, and then click Train.

Account users need IAM access

IBM Cloud services are transitioning from using Cloud Foundry organizations for access control to using Identity and Access management (IAM).

When you add users to your IBM Cloud account, you must now assign IAM Editor access to non-administrative users as well as adding them to a Cloud Foundry organization.

If you previously added users to your IBM Cloud account without assigning IAM access, your users might not be able to provision instances of some services, for example, Visual Recognition. To allow users to provision service instances that use IAM, assign them IAM Editor access.

Week ending 22 June 2018

Data Refinery enhancements to the Filter operation for date and timestamp columns

Use the Filter operation on date and timestamp columns with the following new operators:

  • Is empty
  • Is equal to
  • Is greater than
  • Is greater than equal to
  • Is less than
  • Is less than or equal to
  • Is not empty
  • Is not equal to

Google BigQuery connector

Projects and catalogs now support connections to Google BigQuery, enabling you to store and retrieve data there.

Dashboards support connected data

You can now use a connected data asset in your project as a data source to create a dashboard.

Reuse dashboards for similar data assets

You can now link an existing dashboard with a different data asset, as long as the column names and data types are the same as the original data asset.

Week ending 15 June 2018

Publish trained models to a catalog

You can now publish a trained model from a project into a catalog so that other users in your organization can view the model details and deploy the model.

To deploy a model from a catalog, first add it to a project.

Download files from folder assets

You can now download files that you access from within a folder asset. The IBM Cloud Object Storage connection asset that is associated with the folder asset must include an Access Key and a Secret Key for the download to succeed.

Export and import dashboards

You can now export a dashboard by downloading it as a JSON file to your local system. You can then import that dashboard to a different project by uploading the JSON file on the New dashboard screen. When you open an imported dashboard, you are notified if any of the necessary data assets are missing. After you add the data assets to the project, reopen the dashboard and relink the data assets.

Data Refinery enhancements

Enhancements for handling unstructured text:

  • Use the new Tokenize operation to break up English text into words, sentences, paragraphs, lines, characters, or by regular expression boundaries.
  • Use the new Remove stop words operation to remove English stop words.
  • The Filter operation has new operators that support text and regular expression patterns: Contains, Does not contain, Starts with, Does not start with, Ends with, Does not end with.
  • Use the Sample operations to specify a random sampling step, or automatic or manual stratified sampling based on a step in the flow. Sampling steps from UI operations apply only when the flow is run.

Enhancements to coding operations:

  • Use the new random sampling coding operations (sample_n and sample_frac) to show the result of the sampling in the interactive Refinery tool and apply sampling when the flow is run.
  • Use the mutate coding operation, mutate(provide_new_column = n()), to add a template for counting rows.

Week ending 8 June 2018

Create folder assets in projects and catalogs

You can now create a folder asset based on a path within an IBM Cloud Object Storage system that is accessed through a connection. You can view the files and subfolders that share the path with the folder asset. The files that you can view within the folder asset are not themselves data assets. For example, you can create a folder asset for a path that contains news feeds that are continuously updated. You can preview the contents of files in the folder asset and use the Data Refinery tool to manipulate the data in the files.

Files in folder assets are subject only to policies that operate on the folder asset. policies cannot operate directly on files in folder assets.

Encrypt your IBM Cloud Object Storage instance with your own key

You can now encrypt the Cloud Object Storage instance that you use for projects and catalogs with your own key. You must also have an instance of the IBM Key Project service.

Unicode characters in asset names

Enforced policy chart on the data dashboard in Watson Knowledge Catalog

The Policy enforcements over time section on the data dashboard now shows a line chart for enforced policies. You can select the time span and granularity to display the information on a daily or monthly basis.

Discover data assets from Oracle and Apache Hive

You can discover assets from connections to Oracle and Apache Hive data sources.

Week ending 1 June 2018

Watson Studio

The following new feature is specific to Watson Studio.

Core ML support

After you train your IBM Watson Machine Learning model, you can now download a Core ML (.mlmodel) file to build into your iOS apps.

Week ending 25 May 2018

Watson

The following new features are included in all Watson services.

Data refining

New operators for Filter operation
The Filter operation in the FREQUENTLY USED category now supports these additional operators:

Operator Numeric String Boolean
Contains
Does not contain
Ends with
Does not end with
Starts with
Does not start with

Watson Studio

The following new feature is specific to Watson Studio.

Multiple instances of Watson Studio instances in an IBM Cloud account

You can now provision multiple instances of the Watson Studio services in an IBM Cloud account.

Week ending 18 May 2018

Watson

The following new features are included in all Watson services.

Data refining

Data flow output pane changes
The Data flow output pane is now in display mode by default. If you want to change any of the output details, click Edit Output to put the pane into edit mode. After you save your changes, the pane is returned to display mode.

Enhanced target options for connected data assets
Data Refinery now supports the same target options for connected data assets that it supports for the underlying connections. For example, it supports the same target options for both connected relational data assets and for connections to tables in relational databases. This includes the options for impacting an existing data set (Overwrite, Recreate, Insert, Update, Upsert). As another example, Data Refinery supports the same target options for both file-based connected data assets and for connections to files.

Week ending 11 May 2018

Watson

The following new features are included in all Watson services.

Data refining

Names of data flow run initiators now displayed
When you view the log for a data flow run, you’ll now see the name (instead of a unique ID) of the user who initiated the run at the top of the log.

Template-level command line assistance
The command line has new template-level help. After selecting an operation, simply click the operation name and select a syntax template. Use the template and the content assist to quickly and easily create a customized operation that you can apply to your data.

Project readme file

You can now document your project in a readme file using standard Markdown formatting. The readme is at the bottom of the project Overview page for all new or existing projects that use IBM Cloud Object Storage. Legacy projects won't have readme files.

Week ending 4 May 2018

Watson

The following new features are included in all Watson services.

Data refining

Data flow run initiators added to logs
When you view the log for a data flow run, you’ll now see the unique ID of the user who initiated the run at the top of the log.

Week ending 27 April 2018

Watson

The following new features are included in all Watson services.

Data refining

New Substring operation
The Substring operation in the Text category can create substrings from column values. You simply indicate the starting position within the text and the length of each substring. As with many operations, you can overwrite the current column values or you can create a new column to hold the substrings.

Browsing lots of data flow runs
You can now easily browse a large number of data flow runs on the data flow details page. If there are more runs on the History tab than are currently visible, just click the new Show More button to see more runs.

Watson Studio

The following new feature is specific to Watson Studio.

Project-lib for R save_data function

You can now save data to the object storage associated with your project by using the project-lib library for R.

Watson Knowledge Catalog

The following new feature is specific to Watson Knowledge Catalog.

Easily specify IBM Cloud account emails

When specifying the owner of a business term or defining conditions that require user IDs in the rule builder, start typing the name or email address of a user in your IBM Cloud account. You can then choose the account email from a selection list.

Week ending 20 April 2018

Watson

The following new features are included in all Watson services.

Watson Analytics connector: support for Sydney data center

The Watson Analytics connector now provides support for the Sydney data center. When you create a new target connection to Watson Analytics, you can select AP1-Sydney as the data center in the Connection details section.

Microsoft Azure SQL Database connector: support for secure gateway

The Microsoft Azure SQL Database connector now provides support for the secure gateway. When you create a new connection to Microsoft Azure SQL Database, you can select the Use a secure gateway option in the Connection details section.

Watson Knowledge Catalog

The following new features are specific to Watson Knowledge Catalog.

Business terms restricted to one owner

You can now select only one owner for a business term in the business glossary. One or several names are no longer supported and will be removed. The owner must be an IBM Cloud registered user. If you entered a name or a non-registered email address, these entries will be removed.

View and download asset files from a catalog

You can now view and download files that are associated with assets from a catalog. For example, if you uploaded a data file or a PDF file as a data asset, catalog collaborators can download the file from the asset Overview page.

Week ending 13 April 2018

Watson

The following new features are included in all Watson services.

Changes to top-level menus

To be more consistent with IBM Cloud, some menus and menu items in the header are moved or new:

  • You can now switch your account or organization from your avatar menu.
  • You now access administrative pages from the new Manage menu. The Manage menu also has options to manage your IBM Cloud account.
  • You now access the FAQ, the What's New blog entries, and give feedback on the new Support menu.
  • You now access the Watson Studio and Watson Knowledge Catalog documentation by clicking the Docs button instead of an icon.

Watson Knowledge Catalog

The following new feature is specific to Watson Knowledge Catalog.

Mask sensitive data in columns

You can now protect sensitive data while allowing access to the rest of the data asset. You can create policies that mask sensitive values for a column in a data asset when users view the asset in a catalog or work with the asset in a project.

Week ending 6 April 2018

Watson

The following new features are included in all Watson services.

Data refining

Browsing lots of data assets and connections
You can now browse a large number of data assets and connections. When selecting a data asset or selecting data from a connection, you'll see new Show More buttons at the bottom of the Data Asset tab and the Connections tab. To see more assets or connections than are currently visible, just click this button to get another page of items.

Timestamp support
The Convert column type operation in the FREQUENTLY USED category now supports String to Timestamp conversions and Timestamp to String conversions.

Enhanced date support
When converting a column from String to Date, you no longer need to ensure that the column data is in MM/DD/YYYY or MM-DD-YYYY format. You'll now be prompted to select the current order of the month, day, and year in the date values.

Watson Studio

The following new features are specific to Watson Studio.

IBM Watson Visual Recognition

Enhanced face model GA The enhanced face detection model is now GA! This enhanced model includes reduced bias, increased accuracy of facial detection for age and gender, and tighter age ranges. Existing users of the GA /v3/detect_faces endpoint will not have to do anything, as the enhancements will be automatic. Users of the beta endpoint will need to change their requests to point to the GA endpoint by May 17, 2018, as the beta endpoint will be deprecated and no longer accessible.

Week ending 30 March 2018

Watson

Enhanced target connection support

The following connectors, which supported only source connections in the past, now support target connections too:

  • IBM services
    • Compose for PostgreSQL
    • Informix
  • Third-party services
    • Amazon Redshift
    • Pivotal Greenplum

Watson Studio

Notebook creation with Apache Spark

When you create a notebook using Apache Spark, you can only run the notebook in Spark 2.1. Spark versions 1.6 and 2.0 are no longer available for selection during notebook creation.

Week ending 23 March 2018

Watson Knowledge Catalog

Preview Microsoft Excel documents

You can now see the contents of Microsoft Excel documents that you add to a catalog on the asset’s Overview page.

Data Refinery

Enhanced date support

The Convert column type operation in the FREQUENTLY USED category now supports String to Date and Date to String conversions. When converting a column from String to Date, ensure that the column data is in either MM/DD/YYYY or MM-DD-YYYY format.

Week of 20 March 2018

Watson

New names for the Data Science Experience and Data Catalog services!

The new names better align with new AI features:

  • Data Science Experience is now named Watson Studio.
  • Data Catalog is now named Watson Knowledge Catalog.

Machine learning and AI

Image classification with Visual Recognition

You can now use IBM Watson Visual Recognition within Watson Studio to classify images. Visual Recognition uses deep learning algorithms to analyze images for scenes, objects, faces, and other content. You use the Visual Recognition model builder tool to quickly and easily train and test custom models.

Deep learning

You can now use deep learning techniques to train thousands of models to identify the right combination of data plus hyperparameters that optimize the performance of your neural networks. You can run more experiments faster. You can train deeper networks and explore broader hyperparameters spaces. Watson Machine Learning accelerates this interactive cycle by simplifying the process to train models in parallel with an on-demand GPU compute cluster.

You can use the Experiment Builder tool to define training runs for your experiment and automatically optimize hyperparameters.

You can use the neural network designer tool to create deep learning flows. Design deep models for the following types of data: image (CNN architecture), as well as text and audio data (RNN architecture). The neural network designer supports 31 types of layers. Any architecture that can be designed using the combination of these 31 layers, can be designed by using the flow modeler and then publish it as a training definition file.

Modeler flows

You can now create a machine learning flow, which is a graphical representation of a data model, or a deep learning flow, which is a graphical representation of a neural network design, by using the Flow Editor. Use it to prepare or shape data, train or deploy a model, or transform data and export it back to a database table or file in IBM Cloud Object Storage.

Watson Studio

Create a project with tools specific to your needs

When you create the project, you can now choose the project tile that fits your needs. The tile selection affects the type of assets you can add to the project, the tools you can use, and the IBM Cloud services you need.

You can choose from these tiles when you create a project from the Watson Studio home page:

  • Basic: Add collaborators and data assets.
  • Complete: All tools are available. You can add services as you need them.
  • Data Preparation: Cleanse and shape data.
  • Jupyter notebooks: Analyze data with Jupyter notebooks or RStudio.
  • Experiment Builder: Develop neural networks and test them in deep learning experiments.
  • Modeler: Build, train, test, and deploy machine learning models.
  • Streams Designer: Ingest streaming data.
  • Visual Recognition: Classify images.

If you create a project from the My Projects page, your project has all tools.

After you create the project, you can add or remove tools on the Settings page.

Create dashboards to visualize data without coding

With a Cognos dashboard, you can build sophisticated visualizations of your analytics results, communicate the insights that you've discovered in your data on the dashboard, and then share the dashboard with others.

Customization support for Python environments

You can customize the software configuration of the Python environments which you create.

Watson Knowledge Catalog

Refine catalog data assets

You can now refine data assets that contain relational data after you add them to a project. Projects that you create with Watson Knowledge Catalog include the Data Refinery tool so that you can cleanse and shape data.

Profile documents with unstructured data

Data assets that contain unstructured data, such as Microsoft Word, PDF, HTML, and plain text documents, are automatically profiled by IBM Watson Natural Language Understanding to show the distribution of inferred subject categories, concepts, sentiment, and emotions for the document on the asset’s Profile page. You can also see the profile when you add the asset to a project.

Preview PDF documents

You can now see the contents of PDF documents that you add to a catalog on the asset’s Overview page.

Review and rate assets

You can now review and rate an asset, or read reviews by other users in a catalog. View the asset and go to its Reviews page to read reviews or to add a review and a rating.

Data Refinery

Target file format

If you select a file in a connection as the target for your data flow output, you can now select one of the following formats for that file:

  • AVRO - Apache Avro
  • CSV - Comma-separated values
  • JSON - JavaScript Object Notation
  • PARQ - Apache Parquet

Week ending 9 March 2018

Watson

New style

You'll notice color and font style changes across the Watson services and tools. These changes align with the style of IBM Cloud to provide a more consistent user experience.

New Watson Analytics connector

Projects and catalogs now support connections to IBM Watson Analytics, enabling you to store data there. (The Watson Analytics connector supports target connections only.)

Watson Studio

PixieDust 1.1.18 adds PixieDebugger

PixieDust release 1.1.8 introduces a visual Python debugger for Jupyter Notebooks: PixieDebugger. It is built as a PixieApp, and includes a source editor, local variable inspector, console output, the ability to evaluate Python expressions in the current context, breakpoints management, and a toolbar for controlling code execution. In addition to debugging traditional notebook cells, PixieDebugger also works to debug PixieApps, which is especially useful when troubleshooting issues with routes.

Data Refinery

Basic date and time support

Some Data Refinery operations now support datetime values.

  • Convert column type (support for converting from datetime values only)
  • Remove
  • Rename
  • Sort ascending
  • Sort descending

Watch for more operations to provide this support in the future!

New Substitute operation

The Substitute operation in the FREQUENTLY USED category can obscure sensitive information from view by substituting a random string of characters for the actual data in the column.

Catalogs

Import assets from IBM InfoSphere Information Governance Catalog

You can import assets into a catalog from an Information Governance Catalog archive file. You must have the Watson Knowledge Catalog Professional plan and have the Admin role in the catalog to import Information Governance Catalog assets.

Week ending 2 March 2018

Watson Studio

Apache Spark Service Python 3.5 notebooks now on Anaconda 5.0

The Apache Spark Service upgraded the Anaconda distribution used for Watson Studio notebook environments to Anaconda 5.0. This updated version of Anaconda forces an upgrade to libraries that will change the version for libraries previously installed in the Watson Studio notebook environment. Some libraries updated by this upgrade have changed their APIs, which might cause your existing code to throw warnings or errors.

RStudio and R version upgraded

RStudio in Watson Studio is upgraded to version 1.1.419 and R in RStudio is now version 3.4.3. You might have to update some packages to work with the new R version.

Streams Designer

Event Streams (Source) operator configuration

Until now, if a streams flow stopped and Event Streams producers continued to send messages to the topic, those messages were retained in the Event Streams queue. When the streams flow restarted, it could not go back in time and consume those lost messages.

Now, in the Properties pane of Event Streams (Source operator), you can select the Resume reading check box to start reading in the Event Streams queue from where the streams flow left off.

You can also configure Default Offset where to begin reading in the Event Streams queue when the streams flow runs for the first time, when Resume reading is not selected, or when the resumption offset is lost. You select to start reading from the latest message or from the earliest message.

User-installed Python libraries

In addition to the supported and pre-installed packages, your streams flow might need other packages for specific work. For these cases, you can install Python packages that are managed by the pip package management system. The packages are found at Python Package Index. By default, pip installs the latest version of a package, but you can install other versions.

In Streams Designer, edit the streams flow that will use the package. Click Settings, and then click Environment.

Week ending 23 February 2018

Data Refinery

Snapshot view

You can see what your data looked like at any point in time by simply clicking a step in the data flow. This puts Data Refinery into snapshot view. For example, if you click the data source step, you'll see what your data looked like before you started refining it. You can also click any operation step to see what your data looked like after that operation was applied.

New operation descriptions

Data Refinery provides a description for each operation in the Steps tab. (This replaces the R code that was previously displayed.)

Insert, edit, and delete operations in a data flow

Previously, you could delete the last operation step in a data flow. Beginning this week, you can also insert, edit, and delete any operation step in a data flow.

Cancel a data flow run

You can cancel a data flow run when it's in progress, that is, when its status is Running. To cancel a run, select Cancel from the run's menu on the History tab of the data flow details page.

Insert and update rows in relational database targets

If you select an existing relational database table or view as the target for your data flow output, you have a number of options for impacting the existing data set.

  • Overwrite - Drops the existing data set and recreates it with the rows in the data flow output
  • Truncate - Delete the rows in the existing data set and replace them with the rows in the data flow output
  • Insert Only (Append) - Append all rows of the data flow output to the existing data set
  • Update Only - Update rows in the existing data set with the data flow output; don’t insert any new rows
  • Upsert (Merge) - Update rows in the existing data set and append the rest of the data flow output to it

For the Update Only and Upsert (Merge) options, you'll need to select the columns in the output data set to compare to columns in the existing data set. The output and target data sets must have the same number of columns, and the columns must have the same names and data types in both data sets.

Week ending 16 February 2018

Watson Studio

Environments for notebooks (Beta)

In this beta release of environments, you can select default Anaconda environments with different hardware and software configurations for running Jupyter notebooks. You can have more than one environment in a project and then associate these environments with your notebooks depending on the hardware and software requirements of each notebook.

Policies

View statistics about data assets with personal or restricted information

The Data Dashboard has been extended. You can now check how many data assets contain personal or restricted data. By default, the following classifications are identified: sensitive personal information (SPI), personally identifiable information (PII), or confidential. You can also use your own business terms instead of these classifications.

Choose email addresses from a list in the Rule Builder

When you create a rule in the Rule Builder and need to specify email addresses, start typing and then you can choose from a list of matching email adresses.

Week ending 9 February 2018

Data Refinery

Create, edit, and delete data flow schedules

When you save or run a new data flow, you can add a one-time or repeating schedule for that data flow. You can subsequently edit or delete the schedule from Data Refinery as well.

Scheduled data flow runs are displayed on the Schedule tab of the data flow details page. Past data flow runs are displayed on the History tab of the same page.

Preview source and target data sets from the data flow details page

You view summary information for a data flow by going to the project > Assets tab > Data flows section and clicking the data flow you're interested in. In the Summary section, you can now preview both the source and target data sets.

Watson Studio

Object Storage OpenStack Swift deprecation

When you create a project, use IBM Cloud Object Storage instead of Object Storage OpenStack Swift.

Object Storage OpenStack Swift is no longer available when you create a project if you access Watson Studio from the US-South Dallas region with the dataplatform.ibm.com URL. The Object Storage OpenStack Swift service is available until the end of March, 2018 in the United Kingdom region with the eu-gb.dataplatform.ibm.com URL. Projects with Object Storage OpenStack Swift continue to work.

Easily add Community data sets to a project and notebook

You can add a Community data set to a project by clicking the Add to project button on the data set and selecting a project. Then you can use the Insert to code function for the data set within a notebook.

New Python connector for IBM Cloud Object Storage

You can now use Python connector code in a notebook to load data from and save data to an IBM Cloud Object Storage instance.

PixieDust 1.1.7 is available

PixieDust release 1.1.7 adds support for aggregate value filtering, updates table visualization, improves Brunel rendering, and has some updated icons.

Week ending 2 February 2018

Watson

New Services menu

The Data Services menu is now the Services menu, with new options to add and manage IBM Cloud AI and compute services, as well as data services.

Streams Designer

New canvas design

Check out the new appearance of the Streams Designer canvas! It now has the same look and feel as the Watson Studio common canvas.

Take note of these changes:

  • The bottom tool bar actions (Settings, Run, Save, Metrics) were moved to the top tool bar.

  • The Close button is gone.

New operators

  • Code (in Sources list of operators)

Previously, the Code operator was only a Processing and Analytics type of operator. Now, the Code operator is also available as a Source operator. This operator gives you a convenient way to generate your own sample data or to consume data from an external source.

  • Python Machine Learning (in Processing and Analytics list of operators)

This operator provides a simple way to run Python models of popular frameworks for real time prediction and scoring. The Python ML operator is based on the Code operator. In addition, it can upload the model file objects from Cloud Object Storage and generate the necessary callbacks in the code.

Data Refinery

Save data flow output as a data asset

You can save data flow output as a new data asset or you can replace an existing data asset. By default, data flow output is saved as a new data asset in the project.

To specify that your data flow output be saved as an existing data asset:

  1. From the Data flow output pane, click Change Location.
  2. Select the data asset you want to replace. Note that the target name changes to the name of the existing asset.
  3. Click Save Location.

Change your column selection in the Operation pane

After you choose an operation, you can change the column that you want to apply the operation to. Just click Change Column Selection at the top of the Operation pane, select a new column, and click Save.

New progress indicator

A progress indicator is now displayed when you choose to refine a data set. The indicator provides useful information about what's going on behind the scenes of Data Refinery.

Week ending 26 January 2018

Watson

New Teradata connector

Projects and catalogs now support connections to Teradata, enabling you to access data stored there.

Watson Studio

Any collaborator can leave a project

You can leave a project, regardless of your role in it. Previously, only collaborators with the Admin role could leave a project.

Data Refinery

Data sample size

The name of the source file and the number of rows in the data sample are now displayed at the bottom of Data Refinery. (A data sample is the subset of data that's read from the data source and visible in Data Refinery. It enables you to work quickly and efficiently while building your data flow.)

Preview data sources

When you're selecting data to add to Data Refinery, you can now preview a data source before selecting it. Simply click the eye icon next to the file, table, or view that you want to preview.

Week ending 19 January 2018

Watson

See your current account

If you are a user who can access Watson services in other IBM Cloud accounts because you've been added as a user in those accounts, now you can quickly see which account you are logged into by clicking your profile avatar. The account shows under your user name. You can switch accounts with your avatar menu.

New Dropbox connector

Projects and catalogs now support connections to Dropbox, enabling you to access files stored there. To obtain the application token that's needed to configure a Dropbox connection, follow the instructions in the Dropbox OAuth guide.

Policies

Edit and delete capabilities

You can now edit and delete more policy items:

  • Delete business terms: In Business Glossary, you can now delete business terms that are in draft or archived state.
  • Delete policies: In Policy Manager, you can delete draft or archived policies.
  • Edit rules: You can update rules in published policies if you have the Admin role for the Watson Knowledge Catalog service. The updated rule applies to all other published policies that contain that rule.

Governance Dashboard is renamed to Data Dashboard

The Governance Dashboard is now called Data Dashboard. If you have the necessary permissions, you can see the Data Dashboard by choosing Governance > Data Dashboard.

Watson Knowledge Catalog

Discover assets from PostgreSQL

You can now discover assets from connections to PostgreSQL data sources.

Connections to IBM Cloud Object Storage are on the Settings page

You can now see the connections to your IBM Cloud Object Storage instance on the Settings page of the catalog. The connections no longer appear in the list of catalog assets.

Mark connection assets as private

You can now mark a connection asset as private so that only the connection asset members can see and use the connection.

See policy information when assets are blocked

When an asset is blocked by policies, you now see a message that identifies the policy.

Streams Designer

Use the new Streams Designer tutorials to gain hands-on experience in designing, running, and troubleshooting your stream flows. You can watch videos or follow along the tutorial to see how easy it is to design and deploy a streams flow.

Data Refinery

Larger subsets

To enable you to work quickly and efficiently when creating a data flow, Data Refinery operates on a subset of rows in each data set. Beginning this week, the size of that subset is larger (750 KB). This enables you to see more of your data and use more data for interactive cleansing and shaping operations.

Watson Studio

PixieDust 1.1.6 is available

PixieDust release 1.1.6 updates the Bokeh version and fixes a Bokeh display problem. PixieApps now automatically collapses dropdowns.

Week ending 12 January 2018

Watson Studio

Library to interact with project assets within notebooks

You can use the pre-installed project-lib library in Python notebooks to interact with projects and project assets. Using the project-lib library, you can access project metadata and assets, including files and connections. The library also contains functions that simplify fetching files from the object storage associated with the project. See Use project-lib to interact with projects and project assets.

Dive deeper with enhanced flow editor topics

Lists of the Modeler nodes and SparkML nodes now provide you with more detail about each of the node controls and functions.

Watson Machine Learning topics re-engineered for the way you work

It's a flow thing. We've reworked the order of topics to reflect the way data scientists are using our product. By making use of extensive feedback and leveraging the content on the IBM Cloud we're hoping to make it easier for you to learn as you go.

Watson Knowledge Catalog

Recent catalogs in the Catalogs menu

You can quickly open catalogs that you've accessed recently from the Catalogs menu. Previously, you chose View All Catalogs to go to the Your Catalogs page and then opened the catalog you wanted.

Week ending 5 January 2018

Watson

Important information appears in an announcement bar

If there’s an important product update or great new feature that we think you need to know about, it appears in an announcement bar at the top of the screen. You can easily dismiss announcements, and if you want to read them again, click the notification bell to see the notifications log.

Watson Studio

Invite project collaborators with an email list

It's easier to invite multiple collaborators to a project. You can paste a list of email addresses that are separated by commas into the Invite field, instead of pressing Enter between each email address.

PixieDust 1.1.5 is available

PixieDust now properly supports Python’s string format operator: %. When you define PixieApp views, you can choose to use Markdown syntax instead of HTML.

Watson Knowledge Catalog

Improved navigation for policies

In the Business Glossary, you can use breadcrumbs to quickly jump to previous screens.

In the Policy Manager, you can sort by policy status to find your policies within a category. By default, all published policies are now displayed first.

Parent topic: What's new

Generative AI search and answer
These answers are generated by a large language model in watsonx.ai based on content from the product documentation. Learn more