Getting the most from Discovery

Advantages of using the latest version

Discovery v2 offers the following features and enhancements:

• A project-based experience that supports many different use cases within a single environment.
• Built-in customization tools for adding dictionaries, patterns, and classifiers to help business users build projects that understand the language of their domain.
• Connectors to popular data sources that can quickly access valuable data where it resides.
• Smart Document Understanding that learns from the structure of human-readable documents, such as PDFs.
• Natural language query support across all document types, optimized with machine learning to find targeted answers.
• Advanced search capabilities, such as answer finding, curations, and table retrieval.
• An out-of-the-box contract understanding function that helps you search and interpret legal contracts.
• A full-featured Content Mining application that you can use to conduct in-depth analysis of unstructured text.
• Customizable user interface components that help you to deploy custom applications.

For more information, see Migrating to Discovery v2.

Comparing v1 and v2 features

If you are already familiar with Discovery v1, learn more about how Discovery v2 compares.

Discovery v2 has new features that were previously unavailable. The following table describes feature support in both versions.

Feature support details
This table has row and column headers. The row headers identify features. The column headers identify the different versions of the product. To understand which features are supported by a product version, go to the row that describes the feature, and find the column for the product version that you are interested in.

Feature	Product redesign (v2)	Earlier version (v1)
Use projects to organize your work
Use the Smart Document Understanding (SDU) to annotate your documents
Leverage intuitive user interface tools to add domain-specific artifacts, such as dictionaries and custom machine learning models
Create a content mining project type and then use the built-in Content Mining application to do in-depth data analysis (IBM Cloud Pak for Data, Enterprise, and Premium plans only)
Perform real-time NLP with the Analyze API (IBM Cloud Pak for Data and Enterprise plans only)
Apply a pretrained Smart Document Understanding model to your collection for similar benefits with less effort
Process text from scanned documents or other images
Extract meaning from tables
Get insights from contracts (IBM Cloud Pak for Data, Enterprise, and Premium plans only)
Apply the Part of Speech enrichment to your data
Use the Entity Extraction, Document and Phrase Sentiment Analysis, and Keyword Extraction enrichments
Use the Category classification, Concept tagging, Relation Extraction, Emotion Analysis, and Semantic Role Extraction, Sentiment of Keywords and Entities enrichments, which are available with the Natural Language Understanding service
Build a custom entity type system
Apply Watson Knowledge Studio NLP models to your data
Support for more connectors from a IBM Cloud Pak for Data deployment, including databases, file systems, FileNet P8, and HCL Notes
Some connectors support document-level security from a IBM Cloud Pak for Data deployment
Programmatically configure external data source crawls
Configure the normalization processes of document segmentation and HTML file inclusion or exclusion rules during ingestion
Configure the JSON normalization process during ingestion and after enrichment
Configure dictionary tokenization
Advanced question-answering capabilities, such as returning the exact answer
Discovery Query Language (DQL) API support
Retrieve passages from documents
Perform relevancy training to improve query results
Configure continuous relevancy training
Retrieve tables
Query result deduplication
Identify document similarity in query results
Indicate a preference (bias) in queries
Review query logging and metrics

Limit details

For more information about artifact limits per plan, see the feature documentation:

• Advanced rules model limits
• Classifier limits
• Collection limits
• Dictionary limits
• Document limits
• Entity extractor limits
• Machine Learning model limits
• Pattern limits
• Project limits
• Query limits
• Regular expression limits
• SDU limits

The following limits apply only to Content Mining project types:

• Document classifier limits
• Regular expression pattern limits

To check the current status of the limits and usage for your plan type, you can open the Plan limits and usage page at any time.

1. From the product page header, click the user icon .

   The Usage section shows a short summary.

2. Click View all to see usage information for all of the plan limit categories.

To leave the page, click the web browser back button or the My Projects tab.